Blogs

A good approach might be to post the proposed statement internally, poll the membership and include the results of the poll when publishing the official statement. I recommend permitting the following responses to the poll:

agree
agree with reservations
disagree
do not make an official statement

Ron:

Thanks for raising this important and timely question.

My proposed criteria are that the ASA should either have
1.  Standing -- i.e., a dog in the hunt, or
2.  special expertise.
In addition, there should be
3.  something like consensus within the association.

For example, if the government sought to close a statistical agency, ASA members would lose their jobs (1), and the nation would lose a source of objective information (2).  Statisticians across the board would expect ASA to raise the alarm (3).

The statement on the President's executive order (EO) on immigration met none of those criteria:  1. You asked members to tell their stories of being affected by the EO, presumably because no such stories yet existed (and I haven't heard any since).  2. The EO's purpose was to enhance national security.  Did ASA have data to suggest that the attendant inconveniences to scientific exchange would outweigh its security benefits?  If so, the statement didn't cite them.  3. Was there consensus?  I doubt it, but since nobody asked the members' opinions, how could you know?

Our profession has enormous credibility in its area of expertise.  If we, through our primary professional organization, start engaging in political issues that do not directly affect us, we risk losing our reputation for objectivity.  Let's stick to what we know.

Regards.

Daniel Heitjan
ASA Fellow, 1997
Carrollton, TX  USA

Thanks for starting this discussion!

I think it's entirely appropriate in some cases for the ASA to comment on issues pertaining to scientific freedom. One of the core functions of statistics as a discipline is to provide the epistemological toolset that helps further scientific inquiry. If the freedom to do and report that science in an honest way is threatened, many of our livelihoods will be, too. So for instance I think the carefully worded statement of support for the March for Science and the choice to sign on to the recent AAAS letter about data integrity were both in-bounds, not just from the standpoint that the ASA should reflect our professional commitments as statisticians, but also from the standpoint that this is our guild and it's reasonable for it to speak to protect our industry, too.

I think human rights issues should be considered carefully on a case-by-case basis. Our discipline has several founding fathers who were eugenicists; that is part of our moral heritage. So I don't even know if it's enough to say that it's in-bounds for our professional society to comment on any moral issue with a statistical underpinning, because frankly with this track record I'm not sure that, as a group, we're going to get it right.

Nevertheless: looking through the linked list of statements ASA has signed, I don't think I've found any that I object to, so whether or not I can articulate the line I'd like to see, I trust the current leadership to chart a course that does not cross it. I personally welcomed the statement on holding conferences in states with discriminatory laws, and think that the conference code of conduct framework (which I also endorse) is a good way to think about it. I read it as much narrower than a general statement on human rights -- this is about protecting the membership. The 2014 statement on the rights of the disabled seems to have been argued partly from that vantage point as well.

These statements should be performed infrequently, no more than 5 times a year. Otherwise they will have little impact. Also ASA needs to pay attention to its non-profit and tax-deduction status in making statements. It is not in service to members and benefactors to risk this status because of advocacy positions.

Many thanks to all respondents for your thoughtful points.  

I wanted to insert into the discussion this quick note that the ASA Board based its North Carolina decision on its meeting conduct policy.  Reasonable people may disagree with the position the ASA took, but that was the reason for it.

The challenging part of the "human rights" issue is that many of the arguments are where two perceived rights conflict:  the right to feel comfortable with one's choice of locker rooms vs. the right to not undress in front of a person with the physical characteristics of the opposite sex; the right to control over one's body vs. the right of an unborn person to live; etc.  When the ASA chooses to advocate one "human right" to the disparagement of another, they falsely represent their diverse membership's values without the permission of their members.  I don't want to see the ASA boycotting North Carolina because of the board's particular political views, just as I don't want to see them boycott New York City if they disapprove of its perceived restrictions on gun rights.  Statistical expertise does not guarantee moral expertise.

The last criterion is the one most open to distortion and misuse.  The ASA has members who are strong supporters of both sides in some "human rights" issues (such as the "gay wedding cake" controversy).  The ASA needs to be very cautious in any such area, rather than assuming that their political viewpoint makes the answer obvious.  The board has made some bad decisions in the past, such as endorsing a large spending initiative because it had a small statistical spending component.  You don't want your disagreeing members writing to all congress members saying that the ASA doesn't represent its members.  (I almost did this in one case where the ASA overstepped its mandate.)  We want the ASA to be known as a professional organization, not to be known as a captive of a particular political viewpoint.

This is a well thought out and articulated position.  I agree with the major thrust of the questions that you pose.  I suspect that there is no need to weight the different questions, as reasonable people will view the significance of the "triggering issue" as a part of the equation, e.g., locking up a scientist for research project X versus rescinding a promotion for the same research is a substantively different degree of infringement of scientific freedom.

Where I suspect you might get some debate is on the issue of human/civil rights.  While I personally use classic conservative reasoning to embrace an expansive definition of human/civil rights, this could be an area where the questions could run afoul of religious orthodoxy or contemporary conservative ideology.  But -- you are smart people -- you can navigate those hazards.

Ron,
The selected questions seem appropriate, as does the current decision NOT to develop a formal scoring system - let that evolve as the ASA better discerns the impact of its statements and, perhaps eventually, what topics and by what means it can have greater impact.
I'm very supportive of ASA playing a more active role in these pressing issues and greatly appreciate the efforts of you, Steve, and all of the staff and Board on these matters.

Cheers,
Joel Reynolds

​Ron,
These questions are spot-on!  I am glad that the ASA board has taken this step.  The public and scientific community need to hear from us more often, not less.  However, we need to be careful that when we comment publicly, it has an impact.

I do not think there is a need to scoring these questions.  I believe it right issues will be obvious to us and the rest of the community.  I agree with Denise that we should in some instances engage and support other statistical organization s when the impact is better handled outside our purview.

Thanks for keeping us on our socially- scientifically-responsible toes!
Susan Spruill

Ron

Thank you for starting this important discussion.  I think the questions are exactly the right ones.  I would add two points for consideration -

• would there be a broad consensus on what the ASA might say on an issue (note I am not suggesting unanimity since the ASA is a very diverse community).  The reason I suggest this is because I can envisage issues where everyone agrees it is of relevance to the statistics profession but there is a divergence of views on what the line should be.
• should the ASA be taking a lead on this issue or should the ASA be supporting another national or international statistical society in doing so, as this might change the nature of what the ASA says?  I raise this because not all issues will be nationally specific and some may relate to countries where the ASA has limited understanding or influence.  I would hope that a principle would be for the community of statisticians across the world to work in concert where possible.

I think you are right not to devise a scoring system as not all points will have the same importance for different issues.  It calls for sound judgement and ideally you want to canvass a wide range of views, and in particular ensure that orthodoxies are challenged.  However one of the difficulties is that often to have an impact the ASA needs to make a speedy statement and there is little time to test out alternative views.  To this end it is useful to have a small panel of wise members you can consult quickly.  These members should be prepared to take the occasional risk!  After all the most cautious approach is one of complete invisibility and I would be very sad if you reversed the excellent moves you and Steve have achieved at the ASA (and incidentally Hetan Shah has achieved at the RSS).

My apologies for not getting to JSM this year but I hope it is a great success.

Denise (Lievesley)