How To Tell The President The Facts

A panel discussion at JSM 2003, San Francisco

Co-sponsored by ASA Section on Statistics and the Environment, ASA Section on Statistical Consulting, and the International Environmetric Society

Chair: Peter Guttorp, University of Washington
Panelists: Sally Morton, RAND
                Gerald Van Belle, University of Washington
                Thomas Permutt, US FDA
                Haiganoush Preisler, USDA Forest Service
                Adrián Fernández, Instituto Nacional de Ecología, Mexico

After a brief introduction by the chair, the panelists each gave short presentations.

Sally Morton  described how she told the Secretary of Health and Human Services the facts about Ephedra. Ephedra is a naturopathic substance, and not classified as a drug. Rand performed a meta-analysis of the literature.The result of presenting the study to the Secretary was a ban on advertising the drug, and a warning label reguirement.
Sally's advice:
Gerald Van Belle gave a short (fictitious) presentation intended for the US President on air pollution issues related to ships. There were no numbers on the slides, just in the spoken presentation. He then continued to discuss the concept of "mandated science", i.e. the intersection of science, policy, and values. In the example the science was the effect of air pollution, the policy dealt with the Latino population, and the values had to do with social justice. The US EPA uses the risk assess,emt paradigm: hazard identrification, dose-response, and risk assessment combine to produce risk management. There is a hierarchy consisting of data, information, knowledge and finally wisdom. Wisdom is to know the truth and act accordingly.

Thomas Permutt emphasized that we cannot duck the science quality questions: that is our responsibility. All science is wrong. The question is really how wrong. In order to communicate results we must understand not only the client's language, but also the client's job. It is important to have opinions. There are scientific opinions, and opinions on value and policy. We must be able to distinguish them, but we need to have both kinds. Influential colleagues work in the center of the mandated science Venn diagram.

Haiganoush Preisler advocated whenever possible to say it with pictures. GAM output is graphical, thus easy to explain, even though the model may be complicated both to write down and to fit. Standard errors on a graph really can help. Color graphs are useful, but use at most three colors. Detailed interpretation of  variables are not always very important.

Adrián Fernández described a major policy issue in Mexico, where serious problems with ozone and particulate matter were caused by lax automobile exhaust standards. The 1994 Mexico standards were similar to the 1981 US standards, and Mexico City today is similar to Los ANgeles 30 years ago. The key issue was to deal with new automobiles, but the industry was opposed to stringent standards. In order to get around the opposition, two strategies were used: the ministry proposed a two year exemption to emission inspections in return for adherence to standards later, and the message was sent to the auto manufacturers through newspaper channels (leaks). The responsibility of advocacy is important.

After the presentations, the panelists each got to respond to the others.

Sally: Where do you draw the line for the responsibility of advocacy?

Gerald: Passion is important. We need advocacy with objectivity.

Thomas: The mandated science intersection is a tremendously exciting place to be. It is important to understand the things about which we are objective and those about which we are passionate. To play the game well, we need to be seriously involved.

Haiganoush: Keep it simple. We are well equipped to do this. Not one of us used the word "uncertainty".

Adrián: Without stretching the facts we need to help policymakers by being very clear. Be courageous, take chances, participate in public debates!

The floor was now opened to questions and discussion.

What do you do when decisionmakers want a particular answer from you?

Sally: You need to be precise about what you know.

Thomas: You need to build up a relationship over time. Then you can use it im such situations.

How are you passionate?

Adrian: Be passionate about not letting the government twist the facts. Work with media (leaks), NGOs etc.

What do you do when there is alot of uncertainty, but the issue is still one that needs action?

Gerald: That is when you need wisdom.

Peter: There are different philosophies in different countries. For example, in Sweden one tends to follow the precautionary principle, while in the US actions tend to require proof.

Adrián: One needs to develop cost-effective waus to reduce risk.

Is it our job to lay out options, or to lay out options and say that we prefer one?

Thomas: We nedd to lay out consequences of the options (with uncertainty).

Public comments can sometimes allow non-scientific findings the same weight as scientific comments. How do you reply to the non-scientific statements? Should public opinion be peer reviewed?

Gerald: In this country debated tends to be adversarial. We have a social contract with this country to do science, and have certain responsibilities associated with that. Even when we feel it is a waste of our time.

Would it sometimes help to have different analyses of the same problem?

Thomas: We often operate in that mode. Both sides report to the same advisory committee. This can often make the issue clearer. It is not always adversary–the statisticians on both sides can work together to make clear what there is agreement on.

Adriån: The PM health effect reanalysis issue in the US is a case in point. It is always useful to have more analyses.

Organizational climate has a lot to do with how much a statement is listened to.

Sally: In the Ephedrin case it was very important that the group that made the study did not have ties to alternative medicine.

Gerald: HEI was paid by EPA and the auomotive industry to create a scientific institute acceptable to both. It has an external review process that helps to solidify this acceptability.

Thomas: Yes, the reputation of a government group is very important. That is why it is so important to participate in the policy making.