Economists and statisticians reject contrarian claims about the climate in a blind test

Together with colleagues Tim Ballard, Klaus Oberauer, and Rasmus Benestad, I published an article last week in Global Environmental Change. The basic thrust of the article is readily summarized:

  • We identified representative claims about climatological data made by people who reject the mainstream scientific consensus on climate change.
  • We exchanged the content of those claims, and the labels of the data they were about, into statements about fictitious economic or demographic trends and presented them to expert economists and statisticians to evaluate the accuracy of those claims.
  • In this blind test, the contrarian claims were found to be incompatible with the data and misleading.
  • By contrast, mainstream scientific interpretations of the same data were found to be accurate in the blind test.

In a nutshell, when the political and emotional attributes of climate change are stripped from the data, denialist rhetoric does not pass an expert test whereas the mainstream interpretation of the same data is judged to be accurate.

To illustrate the procedure, our experts would be presented with a claim such as that shown in the figure below:

This claim would be accompanied by a figure with the actual data, shown below:

Readers who are familiar with the climatological literature may recognize the data above as representing the mass balance of glaciers around the world. It is pretty obvious that glaciers worldwide are retreating although there are a few exceptions. It is therefore perhaps unsurprising that the experts judged the contrarian statements to be misleading and incompatible with the actual data, especially in comparison to the mainstream interpretations of the same data, which would of course state the obvious fact that most glaciers are shrinking (in this case, presented as “rural populations shrinking”).

We combined the expert judgments into a “correctness score” for each claim. The figure below compares the overall distribution of scores across all scenarios and participants, separately for the mainstream and denialist statements:

Any positive number indicates expert endorsement of a claim about the data and any negative number represents expert rejection. The data leave little room for ambiguity: the vast majority of judgments endorse mainstream statements and reject the denialist interpretations.

There are some interesting exceptions, which arise because the underlying physics is lost during the translation into economic terms. For instance,  cherry-picking of a very short trend that characterizes much contrarian discourse (“sea levels haven’t risen in 6 weeks!!!”) becomes less recognizable as being misleading when it is applied to the same data labeled as an economic indicator (e.g., balance of trade). When couched in economic terms, the scientifically appropriate emphasis of long-term trends becomes less imperative than with the original climatological context.

At one level, our results are entirely unsurprising: In light of the overwhelming scientific consensus on climate change, most dissenting opinions are merely political and rhetorical tools aimed at trying to forestall mitigative action. The efficacy of those talking points is measured in political not scientific terms, and no measurable positive scientific contribution has ever arisen out of denial. Indeed, the few peer-reviewed articles that are authored by contrarians are by and large flawed.

Our data do add some important novelty, however: Previous judgments about contrarian attempts to create a parallel interpretation of reality were mainly made by climate scientists, who could be perceived as biased in favor of the dominant view of their discipline. In our present study, in contrast, the same negative judgments were made by experts who were unaware of what data they were considering. This rules out the possibility that the experts participating in our study may have been driven by any extraneous considerations in their adjudication.

One strong implication of our study is that denialist claims do not deserve the same amount of attention as scientific statements. Giving them equal weight in media discourse does the public a disservice by denying it the right to be adequately informed about the risks it is facing. Media “balance”, in other words, can be a pernicious form of bias.