Some changes at Social Science One

Social Science One is now housed at Harvard’s Institute for Quantitative Social Science and is supported by a small endowment, which gives us a little elbow room. Unrelatedly, Nate Persily, after two years of devoting his life to this enterprise and all of us in the research community, is stepping down as co-chair (see his statement below). You may also have seen the announcement about our new Facebook 2020 US Election Study and updates to the “Condor” URLs dataset (now covering 46 countries and including more than 17 trillion differentially private cell values), as well as our more general efforts on differential privacy through our new collaboration with Microsoft and our Sloan Foundation grant for a community effort known as OpenDP.

With these developments in process, we have been eliciting ideas from many people for restructuring Social Science One, its committees, leadership, and projects, all toward the goal of helping to maximize the flow of data and expertise among academia, industry, and government. In this process, we have been talking to scholars associated with Social Science One, and widening concentric circles of others. If you have any ideas, I’d love to hear from you via email, phone, v-chat, or any other means. (I wish I could also encourage you to stop by my office, but I also wish I could stop by my office!)

To help with this process, I’ve negotiated more specific agreements with Facebook (as only one of the companies we will work with) about what we will have the authority to do and what we can offer.  These will involve our influence over the creation and distribution of new datasets, better computer infrastructure, improved software and hardware to analyze the data, deeper contacts with relevant parts of the company, and some input into parts of Facebook decision making. We will also be able to offer early access to new datasets, including for publication, for those involved in Social Science One. There is more to it than this, and more to arrange, but here are some details on what we have arranged so far. Obviously, experience with Facebook indicates that we have to see what happens in practice. At the same time, we have amassed lots of experience building different types of industry-academic partnerships, under different models, and so we hope to be able to extend data sharing arrangements for academics with other companies and organizations.

On organizational directions, I think it makes sense for Social Science One leadership to focus on negotiating industry-academic partnerships and supporting the organization, and to devolve any decision making power we might have over data access to subsets of committee members (such as to constitute the “commission” of academics to decide on access to Facebook data under our original partnership model). Another issue is the leadership structure, which could be a chair, co-chairs, an executive committee, or something else. We need to consider how we might revamp our committee structure to better convey publicly what we are doing, and to ensure inclusive access to a diverse array of academics from around the world. One suggestion has been to switch to geographically organized committee structures (and by specific project when needed), rather than substantive areas of interest, or to one diverse group without divisions. Alternatively, maybe we should follow those in public health who regard it as inappropriate to extract data from a developing country without involving local academics, and so another option is to actively foster more vibrant cross-national collaborative research when we are able to make data available.

Many other issues need consideration, and we have many more discussions to come and work to do.  But right now, I know I speak for the whole academic community in expressing my deepest thanks to my distinguished collaborator and co-chair, Nate Persily, for all he has contributed to the social sciences, and for all our work together on this project. Nate worked tirelessly for the academic community on Social Science One since its inception. It is three hours earlier in Palo Alto than Cambridge, but I would often receive texts from him before I woke up and we regularly talked late into the night. The fact that he lives in Silicon Valley near the tech companies was convenient for me and others but it also meant that, in addition to endless electronic communications, he had numerous in person meetings to attend. And, as if that wasn’t enough, Nate took on a diplomat-in-chief role at Social Science One, traveling all over the world to stay in close touch with the disparate parts of the academic community and regulators -- trying to explain the latest inexplicable turn of events with Facebook to everyone (I came to believe that his average location over the last two years was somewhere near the center of the earth).  And all the while, Nate has remained the consummate scholar -- a political scientist, lawyer, and policy analyst and participant -- deeply involved in the legal and policy process.  While helping to run Social Science One, he ensured the efficient and fair operation of democracy in several US states through his redistricting and expert witness work, was commissioner on the Kofi Annan Commission on Elections and Democracy in the Digital Age, and has contributed important scholarship to democracy, elections, voting, and what he calls the “law of democracy”.  And now, as he explains below, we are about to be indebted to him for keeping American elections safe in the 2020 US election. Thanks from us all.

 

Here’s Nate’s statement in full:

I am writing today to announce my resignation as co-chair of Social Science One. It has been a great honor working with Gary King for the past two years, meeting with so many fantastic social scientists around the world to try to bring transparency and to provide independent academic access to social media data.  We have learned many lessons over that period about how difficult it is to dislodge data from social media companies for independent academic analysis. I strongly believe that our experience, difficult as it has been at times, was necessary to craft any future pathway for academic partnerships with technology companies. I will continue to work on these issues, especially promoting government regulation to ensure a privacy-protected pathway for academic access to corporate data, in my role as Co-Director of the Stanford Cyber Policy Center. However, in the short term I need to dedicate my energy to the Stanford-MIT Healthy Elections Project (healthyelections.org) to do what I can to help ensure the 2020 U.S. election is safe, fair, and legitimate.