Why We Need a New Theory of Government

Abstract

In the 1970s I was among a group of scholars endlessly debating theories of the state. Others in the discussion were my recent predecessor as APSA President, Theda Skocpol, and my immediate successor, Ira Katznelson. What intrigued us was a vast literature, grounded in neo-Marxism and covering huge swaths of history and geography. Nearly all the important books and articles were by sociologists and historians, but with Structure and Change in Economic History, my then-colleague, economist Douglass North, transformed the debate by using economic models of transaction costs and property rights to model the state's role in economic prosperity over time. Most political scientists now acknowledge the importance of this perspective, but it nonetheless helped precipitate twenty years of divergence between historical and new economic institutionalists.Margaret Levi is Jere L. Bacharach Professor of International Studies in the Department of Political Science, University of Washington, Seattle (mlevi@u.washington.edu). Among her books are the single-authored Consent, Dissent and Patriotism (1997) and Of Rule and Revenue (1988) and the co-authored Cooperation Without Trust? (2005) and Analytic Narratives (1998). Many people offered me comments. I did not always take their advice, but I am grateful to Amit Ahuja, Paloma Aquilar, Marcelo Bergman, Maureen Eger, Ann Gryzmala-Busse, Bea Kelleigh, Bob Kaplan, Edgar Kiser, Victor Lapuente-Giné, Michael Lipsky, Kenneth Kollman, José-Maria Maravall, Peter May, Leonardo Morlino, Steve Pfaff, Kate Pflaumer, Frances Fox Piven, Christoph Pohlmann, Nancy Rosenblum, Jean-Laurent Rosenthal, Bo Rothstein, Susan Stokes, Katherine Stovel, Joan Tronto, and Ashutosh Varshney. My greatest debt is to the two graduate students who read and commented on several drafts as well as located the materials I needed to write this presentation: John Ahlquist and Audrey Sacks.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 149 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.