I appreciate Ron' s request for our input. From my perspective, the answer is clear. What binds members of the ASA is the desire to practice and propagate proper statistical science. Thus I would expect the organization that represents us to publicly object to the misuse or misapplication of statistical methods wherever it occurs, and to advocate and promote the proper use of such methods whenever it can. I would also expect the ASA to be intimately involved in proper education of statistics in our schools and to speak out if the profession or the ASA itself is maligned or misrepresented.
If I want an organization to be a public voice for me for human rights or even scientific freedom, I should join any number of other civic, social or political organizations which fill that aim. Even bullet # 4 in the ASA Board's proposed criteria for public advocacy leaves the door wide open for the ASA to be my political voice, based on an "expectation" that I and other members would desire "involvement" in an issue based on the "judgement" of the Board. Statisticians are a diverse lot, likely to be on both sides of any issue. I don't want the ASA to take my side or any side - that's not what I pay dues for; there are other organizations I should join instead if that were my goal.
To those of you who wish to have the ASA publicly express views on important 'moral' or political issues, I would likewise suggest you instead join other organizations whose goal it is to represent your particular view. You should make your case as best you can to others without asking the ASA (and by proxy, me) to take sides.
I recommend that the ASA consider four fundamental principles in its decision to issue an official ASA statement:
Any official ASA pronouncement should adhere to its tagline, i.e. it should: 'Promote the practice and profession of statistics".
Any public or official statement or request by any Officer of the ASA should have Board approval and endorsement, including blogs or any informal requests for support which use a title of an office or officer of the ASA, e.g. Science Policy Adviser. Whether formal or informal, attaching an Officer or Office to a statement or blog puts the ASA imprimatur on it.
All statements should reflect the unique view of statisticians, statistical science and statistical thinking and not scientists in general, academicians in general, political parties, or any other group or 'genre'.
The ASA should always advocate for open access to data and models where possible, or to a trusted 3rd party where necessary, especially in the public policy arena. Without the ability to evaluate the underlying analysis and data, it is impossible to verify or validate the conclusions.
Here are examples of what I see are important societal issues which the ASA and the Office of Science Policy could focus on which fit within my criteria above:
For policy questions, form a blue ribbon panel of volunteers of distinguished statisticians to carefully and objectively acquire, then assess, the statistical methods and data used to evaluate various government programs. Example: the efficacy of poverty programs. What measures of effectiveness are used? I know there are some interesting studies about these programs, including economic mobility. Are the statistics well done? Is the data collected upon which the assessments are made reliable or unbiased? What priors are used? This could inform policy makers from a statistical standpoint. (I stress that evaluating the underlying analysis is a statistical question, whereas funding these programs at some level or assessing their cost/benefit is a political question.) The same could be done for any number of policy questions: anthropogenic global warming, veterans care, EPA evaluation of health effects, Medicare, etc. What are the data and statistical methodologies behind the policies?
If one doesn't want to veer into policy, what about a statement and study of the consequences of non-publication of non-significant results in the social sciences? What stand does/should the ASA take and what are their proscriptions? I've seen a scattered paper here or the on the topic, but the ASA has not to my knowledge taken any stand on the appropriateness of the editorial judgments made for journals. What about a clear statement of acceptable data analysis practices in refereed journals? Given that many recent studies have shown that many so-called "results" cannot be replicated, what is the ASA's stand on funding for such studies?
Another example is clinical trials. What does the ASA say about the proper methods and evaluation of trials? How about the reliability of DNA evaluations?
I could think of many other examples where society would benefit from experts in statistics weighing in on the appropriateness of the data and statistics behind important decisions based on empirical information.
On the other hand, the ASA public statements are focused too much on federal budgets and legislation and politics in my opinion. Perhaps that is because the ASA is located in Washington D.C., a town which lives and breathes politics. Here are examples which fall outside what I see the proper role and function of the ASA and hence the ASA should have no public position on them:
Federal budgets for any program. These decisions are political, involving allocation of resources and do not depend on good statistical methodology (although they may be informed by studies I suggested in #1 above). This includes sacred cows such as the census or NIH.
Any legislation on civil rights, human rights, social justice, "sustainable programs", etc. Important issues perhaps, but properly included and evaluated in the political arena, and not an issue of proper statistical methodology.
Any city, state or federal legislation.
To summarize, the ASA should stick to statistics, data access and open analysis, and let others debate and comment on the moral, cultural and political issues of the day. To do otherwise would be to not represent the fundamental purpose of the ASA and therefore to not represent the values of all of its members.