Statistics and Public Policy: A New ASA Journal

David Banks, Duke University; Sharon Lohr, Westat; Dan McCaffrey, ETS; and Sally Morton, University of Pittsburgh

The ASA will soon launch a new journal—Statistics and Public Policy (SPP)—with the goals to publish articles that apply good statistical practice to data sets that may inform the public and decisionmakers about matters of policy and provide a forum for discussion of statistical issues associated with evidence-based public policy and planning. Unlike essentially all other ASA journals, SPP will have no requirement for methodological novelty. Of course, novel methods are welcome, provided they apply to public policy issues and applications. The scope for content can be at any level, from city management to national or international governance.

The new journal will be entirely electronic. Taylor & Francis, the publisher for all other wholly owned ASA journals, will provide copyediting and editorial support. Also, the new journal will be open access, available to all free of charge without a subscription or a charge per article. Open access for the journal furthers the ASA mission to encourage the use of sound statistics by elected officials and public administrators.

All papers submitted to SPP will undergo a rigorous review process. The review will be overseen by one of the four editors (David Banks, Sharon Lohr, Dan McCaffrey, and Sally Morton) and directed by an associate editor (Georgiy Bobashev, Alicia Carriquiry, Ron Fricker, Alan Karr, Mary Elizabeth Landrum, Mike Larsen, Denise Lievesley, David Marker, Jasjeet Sekhon, Bruce Spencer, and Andrew Thomas).

In addition to being open access, Taylor & Francis has agreed to provide moderated comment capability so readers can discuss an article and authors can respond to readers’ questions. This functionality will not be available immediately, but the editors are optimistic that it will be added promptly.

Also, SPP will have a reproducibility certification. Reproducibility is weaker than replication; it means that if the author’s code is run over the author’s data, then the outcomes described in the paper are actually obtained. Reproducibility studies are problematic for many journals—often the statistical analysts do not own the data, or the software is proprietary, and they cannot get permission to share these with the editors. In the public policy sphere, however, most data are open access and most software is commercial or free. The editors will strongly encourage (but not require) authors to submit their work to a reproducibility test.

Open access will be a great benefit for SPP, but it means there is no subscription revenue to cover the costs of the technical support provided by Taylor & Francis. The ASA is exploring methods for implementing the new open-access model. The current plan is that authors may need to pay a fee if their paper is accepted for publication. The ASA Board is sensitive to the fact that this is a departure from previous ASA practice and hopes the membership will agree that, on balance, the positive outcome of world access to content outweighs the relatively small fee. The ASA intends to ease into this payment model, and authors who have good articles but no external support may, upon request, receive a subsidy.