Evolutionary Theory and the Dynamics of Institutional Change ∗

How do we explain the dynamic processes that underlie institutional change? This is a crucial — and still unresolved — question for the social sciences. Sociologists have tended to analyze institutional change in terms of agency, focusing on the key role of entrepreneurial actors in shaping institutional outcomes (Fligstein and McAdam, 2011) . Rational choice scholars in political science and political economy have more and more sought to use game theory to explain how institutional change occurs, focusing in particular on institutions as equilibria that arise in infinitely iterated games (Greif, 1994; Greif et al., 1994; Milgrom et al., 1990). Historical institutionalists have sought increasingly to capture the specific mechanisms that guide institutional change over time (Hacker, 2004; Mahoney and Thelen, 2009; Thelen, 2004). Each of these approaches faces its own particular difficulties. Sociologists’ emphasis on institutional entrepreneurialism helps them to capture the genuine creativity that actors can exercise in creating, recreating or extending rules. However, just because of this focus, sociologists have found it hard to identify possible constraints on entrepreneurialism that would really help build a proper theory. Rational choice scholars’ concern with equilibria makes it very difficult to identify processes of endogenous change (a game theoretic equilibrium is, by definition, self-reinforcing) and furthermore makes it difficult for them to speak convincingly to the kinds of new institutions that might come into being when an old equilibrium breaks down. Historical institutionalists continue to face difficulties in identifying the relationship between major conjunctural change, and processes of gradual interstitial change. It is, as always, far easier to identify imperfections in existing work than to propose useful ways in which they might be fixed. Even so, we seek in this article, — perhaps with more ambition than good sense — to provide a new account of institutional change that at least helps to rectify some of these flaws. To do this, we draw upon an extensive literature in evolutionary theory. An evolutionary account at a minimum requires both a mechanism of transmittable variation, which generates units with transmittable differences, and a mechanism of selection, which determines the ∗We are grateful to Mark Blyth, Ellen Immergut, Elinor Ostrom, Sven Steinmo, and Kathleen Thelen for their comments on an earlier version of this paper. Henry Farrell gratefully acknowledges the support of the Woodrow Wilson International Center for Scholars.