We talked to Rohit Naimpally, research and policy manager at J-PAL North America, about the work that they are doing in administrative data. This conversation has been lightly edited for publication.
Monica King: Could you talk a little about J-PAL North America and how it got started?
Rohit Naimpally: J-PAL is primarily known for our international work supporting and running randomized evaluations. Our North America offices started in 2013 with initial funding from the Alfred P. Sloan Foundation and the Laura and John Arnold Foundation. Over time we realized that there are a lot of commonalities in the research that we do internationally and in North America, but there are some things that are pretty unique to the research done in North America. One of the things is the availability and use of administrative data here. We learned that all the resources we created for our researchers and affiliates in our network have typically focused on research done elsewhere where there hasn’t been this wealth of administrative data.
Tell me more about J-PAL North America’s work with administrative data.
With J-PAL North America, we’ve done two main things in administrative data.
One is a guide to using administrative data for randomized evaluation. This is a 40-page guide that we built that came out last year. Basically it guides people on the specific needs of administrative data and how they can be applied in the context of randomized evaluation. Although it’s targeted very much toward our researchers, we think it’s more broadly applicable to people using administrative data in general. The guide goes through a number of things such as data security, data access, data flow, how to be compliant with various requirements.
The second thing we’ve done is building out our catalog of administrative datasets. The catalog was started with datasets that have been used in J-PAL North America evaluations. With each dataset we looked at a variety of things – where the data are available, whether or not it’s free, what topics are covered, how you can gain access. The idea is to build out this catalog or library of various administrative datasets that are used in social sciences research and economics research.
What are you looking forward to the most in the coming new year?
We’re looking forward to growing our catalog. As part of our plan, we also hope to have more examples of Data Use Agreements signed in the past, examples of informed consent, and provide general guidance and advice on the data sharing process.
In terms of what our policy team does, we’re like to continue to support efforts in the administrative data space. Our executive director, Mary Ann Bates, provided testimony to the Commission on Evidence-Based Policymaking, and it was specifically related to administrative data and the values of linking datasets. We also generally advocate for the value of administrative data in our policy work. We have this publication on important research results and policy-relevant results that came from having more access to administrative data. While we advocate for randomized evaluations, at North America we routinely advocate for the use of administrative data because of the low-cost way to answer important questions.
What are some ways for people to get involved with your efforts at J-PAL?
We have J-PAL mailings and newsletters that people can sign up for. If people want to get involved beyond that, we are also hiring right now. We’re also constantly looking for feedback on our resources. If people have suggestions for datasets we could include in our catalog, we’d love to hear that.
J-PAL North America’s guides and resources on administrative data are available here.