Over the past decade, a trend toward more careful and explicit “causal inference” has spread throughout the political science discipline. This movement – which puts a premium on research designs that can identify empirical evidence that is consistent with causally-based arguments – emerged out of the political methodology literature, and has been applied to various questions in American Politics, Comparative Politics, and International Relations. Such designs include (now-standard) approaches like difference-in-differences, regression discontinuities, and instrumental variables, among others.
One area where the causal inference (CI) movement has not made much headway is in the study of American Political Development (APD). APD has historically operated on a parallel track from modern social-science approaches to American Politics. Some of this is a function of the nature of APD’s emergence – for example, historically-minded political scientists did not believe that the dominant behaviorally-based studies of the 1970s fully appreciated or incorporated the study of history and institutions, and hence set out to create an alternative path built around historical sociological analysis. Nevertheless, APD scholars seek to make causal arguments, albeit in the form of more traditional narrative-based approaches.
We thus envisioned a conference where proponents of CI and APD can come together and discuss how scholarship and approaches in one area can complement scholarship and approaches in the other area. For example, CI scholars could provide advice on how particular techniques and methodologies could help APD scholars develop cleaner and deeper causal arguments. In additional, APD scholars can help causal inference scholars to identify cases – and empirical contexts – that would be candidates for the application of standard CI designs.
We anticipate that the conference will address some or all of the following general questions:
(1) What CI can do for APD?
- CI scholars can help APD scholars understand how exogenous shocks lead to developmental/historical change. So, perhaps, think about the inferential approach in APD stemming from exogenous change specifically. This would be a “branch” of APD that could be called something like “well-identified” causality.
(2) How can APD provide for (better) CI?
- APD scholars can help CI scholars in identifying exogenous shocks. APD scholars possess tremendous raw knowledge of history (and historical events), which can be a great asset in assembling cases for further study.
(3) What to do about issues of “complicated causality?”
- There are lots of issues in the traditional APD literature (time, sequence, general equilibria, path dependency, contingency, policy feedback) that make CI either more difficult or less useful (less likely to generalize, be externally valid, etc.). What should we do in these cases? Are there “best standards” for situations when “clean” or generalizable inference is not possible or practical?
(4) How do various CI methods & measurement techniques apply to APD specifically?
- Here, we might think of a “survey” of different approaches and techniques (difference-in-differences, regression discontinuity, text analysis, ideal point estimation, etc.). We might also discuss challenges in current “normal science” issues like replication, pre-registration, data transparency, etc., and how they apply to APD work.
Jeffery A. Jenkins
University of Southern California
Charles Stewart III
Massachusetts Institute of Technology
Conference to be held at the University of Southern California University Club‘s Scriptorium room.
January 10, 2019 – Day One
8 to 8:55 am: Breakfast
9 to 10:15 am: Panel 1
David Bateman (Cornell): “A Developmental Approach to Historical Causal Inference”
Corrine McConnaughy (George Washington): “The Inferential Opportunity of Specificity”
Discussant: Boris Heersink (Fordham)
10:15 am to 10:30 am: coffee break
10:30 am to 11:45 am: Panel 2
Sanford Gordon (NYU): “Causes, Theories, and the Past in Political Science”
Sean Gailmard (UC-Berkeley): “Formal Modeling in the Study of American Political Development”
Discussant: Kris Kanthak (Pitt)
11:45 am to 1 pm: Lunch
1 pm to 2:15 pm: Panel 3
Sara Chatfield (Denver): “Causal Inference and American Political Development: Contrasts and Complementarities”
Anna Harvey (NYU) & Gregory DeAngelo (Claremont): “Applying Regression Discontinuity Designs to American Political Development”
Discussant: Jon Rogowski (Harvard)
2:30 pm to 4:00 pm Book Talk
Avidit Acharya (Stanford) – Deep Roots : How Slavery Still Shapes Southern Politics
Discussant: Nolan McCarty (Princeton)
January 11, 2019 – Day Two
8 to 9:00 am: Breakfast
9 to 10:15 am: Panel 4
Gregory J. Wawro (Columbia): “Natural Experiments, Causality, and Historical Analysis”
Christian Grose (USC) & Abby Wood (USC) “Causal Inference and Political Reform”
Discussant: Joel Sievert (Texas Tech)
10:15 am to 10:30 am: coffee break
10:30 am to 11:45 am: Panel 5
Sarah Binder (George Washington, Brookings): “How we (should?) study Congress and history”
Joshua D. Clinton (Vanderbilt): “Causal Inference, Agenda Setting, and Roll Calls”
Shigeo Hirano (Columbia): “Civil Service and Local Government Services”
Discussants: Chris Tausanovitch (UCLA) & Jessica Trounstine (UC-Merced)
12:30 pm to 1:45 pm: Lunch
updated Jan 9th
Papers are available only to symposium participants when available, they will go online Monday, January 7.
(Please refer to the password sent to participants via email.)
Lodging & Travel
PLEASE NOTE: Airfare will be booked by out-of-town guests, reimbursed after the event via the The Society for Political Methodology. Click here for reimbursement directions (please refer to the password sent to participants via email).
Out-of-town guests will be staying at USC Hotel (formerly the USC Radisson) located at 3540 South Figueroa Street, Los Angeles, CA 90007. The map below is the correct placement, but google has not yet been updated with new contact numbers and website. Use the link here for information about the “new” hotel.
Refer to the email from Anne Johnson for confirmation numbers, and you can contact the USC Hotel with any other questions 833-226-65872.