25 July 2014

Sack the Science Advisor

So you don't like the advice that your expert advisor is giving? So then how about sacking the advisor? Even better yet, how about getting rid of the entire advisory mechanism?

That is the advice that Greenpeace and other NGOs are giving to European Commission president-elect Jean-Claude Junker. If it sounds a bit like something out of the Richard Nixon playbook, well, it is.

I defend the EU science advisor and the broader structure over at the Guardian. Read it here.

For a deeper dive into science advice in government, see this volume by James Wilsdon and Rob Doubleday. My chapter in that volume on the role of the science advisor can also be found here in PDF.


  1. I'm disappointed. I thought you were recommending sacking John Holdren which would have been a waste of time because even if he's always wrong, he give the administration the answer it wants.

  2. Question is Roger- would you support sacking the hapless Holdren?

  3. Off-topic, but: Junker. What an interesting name! I wonder if his family still has the family crest from the Crusades...

  4. Surely the answer is to insist that scientific advisers contain both the main opposition points (honestly researched, not straw men), and the consequences of following their advice, every time their opinion is asked.

    This is how decent lawyers work. They lay out their case, but they also lay out the weaknesses and consequences of the opinion. Not to mention the actual cost -- not some hopeful "best case" cost.

    This is how we expect our doctors to work. They suggest a procedure/course of medication etc, but ethically they need to warn of likely consequences, or that the procedure is considered risky or controversial.

    So why is a science adviser not expected to work the same way. Why is Holdren, say, allowed to make his case without a true and accurate examination of its weaknesses? Why is he not obliged to point out that his track record of prediction is poor, to say the least? Why is he allowed to make "best case" costings, without pointing out that well-qualified people think his case is excessively optimistic?

  5. The majority of European scientists would agree with her. Greenpeace is increasingly from the school of don't bother me with the facts i know how we feel about this issue. I doubt that Holdren and some of his radical thinking would get that kind of support she got.

  6. Roger, you end your comment at the Guardian with an appeal to successful science advice, how would you define it?

  7. Nixon playbook?! You mean the Obama playbook of getting rid of all those inspector generals who refused to whitewash the administration corruption they uncovered?

  8. The “art” of science or technical advisory to a politician is doubly disingenuous, fraught with pitfalls, raising obvious ethical questions. First, a science advisor is selected because his world view may fit with the politician's and he would support the politician. To be fair, there are similar issues in the private sector. “Independent” consultants or advisers are hired to influence opinion of, e.g., banks, rule makers, public institutions, etc.; however they are paid megabuck fees, and you would have look far and wide to find an independent consultant/adviser giving truly independent opinions. A better way would be for the consultant to work “double blind” - without the consultant being identified and without the hiring party knowing who the consultant is … until after his work is done.

    Secondly the politician obviously selects and uses his science adviser(s) as an “appeal to authority” to control the story line (think about Stephen Chu and John Holdren standing beside the president who proudly claims them as their Nobel prize winning (or whatever) expert advisor. This does several things: first, it introduces a one sided intentional bias and is playing tricks on the audience to be influenced or controlled; second, science advisors are photo-op props (a kind of bobble-head). It is a well known propaganda technique (See Norman Davies – “5 rules of propaganda” in “Europe – a History, Harper Perennial Books, 1998: pp 500-501). Yes, the media are biased historically and today, one way or another, ideologically and directly or indirectly influenced by the politician, sometime to the point of the media running cover for the politician to aid and abet in manipulating public opinion. It doesn’t matter if it is patently obvious that tricks are going on; the media still gives air time to support the storyline they favor.

    The climate area is similar to medical research on hot issues. Examples include phthalate ester plasticizers for PVC plastic and bisphenol A (known as bis A) a raw material for polycarbonate plastic and epoxy resins. Since these are “hot” research areas in the public eye it is tempting easy to knowingly or unknowingly introduce biases in conducting the research (which might include animal research, epidemiological panel studies, etc. Researchers (not all) obtain grant money to study the problem and they are highly subject to selection bias to get results they can publish; they are not going to be able to publish negative results. Fracking chemicals is another visible target today. Climate science is even more difficult since the “research” (my quotations) leading public policy is in the form of output of deterministic parameterized computer models. An example of a “consultant” manipulating the outcome is Nick Stern’s “Stern Review” which was criticized widely including by highly regarded economists participating in the field, i.e., Wm Nordhaus and Richard Tol who agreed that Stern a) cherry picked information to give the answer he wanted, b) Stern’s use of unsupportable high discount rates to make the benefit / cost analysis more favorable.