This volume is a successor to James Mahoney and Dietrich Rueschemeyer’s 2003 Comparative Historical Analysis in the Social Sciences, with entirely new content. Like its predecessor, the book focuses on the substantive, theoretical, and methodological contributions of comparative historical analysis (CHA). Mahoney and Thelen define CHA as the study of large scale, complex, important, and enduring outcomes through deep case-based research that pays attention to processes and the temporal dimension of politics. They argue that these attributes enable CHA to improve our understanding of politics in ways that complement statistical, experimental, and quasi-experimental approaches.

In particular, the editors argue that CHA counteracts three dangers evident in the recent focus on social science experiments (pp. 8–11). First, CHA addresses important issues that are ethically or financially difficult to study in experiments. Second, CHA’s focus on slow-moving structures balances the focus in experiments on micro factors like information that are easily manipulated. Third, CHA focuses on theory-generation as well as theory testing.

One limitation of the introduction and of several other chapters is that they over-emphasize forms of path dependence that involve increasing returns and institutional lock-in. This neglects self-eroding processes and reactive sequences through which institutions are weakened or even reversed, which Mahoney and Tulia Faletti discuss in a later chapter (pp. 220–223).

The substantive section of the book includes chapters by Stephan Haggard on the developmental state literature, Jane Gingrich on the research program that resulted from Gost Esping-Andersen’s Three Worlds of Welfare Capitalism, and Steven Levitsky and Lucan Way on CHA research on authoritarian durability. Each chapter constitutes an excellent literature review that will prove useful in graduate courses on comparative politics.

The third section of the book focuses on the theoretical contributions of CHA. Here, Paul Pierson writes on power and path dependence, arguing that pluralists have focused too exclusively on overt political conflict. Pierson maintains that social scientists are in a better position than ever before to measure subtler dimensions of power, including agenda-setting, the anticipated reactions of powerful actors, and ideational power (he could have added that computer-assisted analysis of political discourse is a promising means of studying such topics). Pierson emphasizes positive feedback and the material resources through which the powerful become more powerful (p. 134), while he is comparatively neglectful of how changes to discourse and legitimacy can undermine even materially endowed actors. Still, Pierson’s chapter convincingly argues that much of politics is an effort not just to win the current battle but to “institutionalize advantage.”

Pierson’s focus on the intended consequences that powerful actors embed in institutions is complemented by a chapter by Pierson, Thelen, and Jacob Hacker on “drift” and “conversion” as modes of institutional change. Drift happens when the social context changes in ways that modify the outcomes of institutional rules, and it may or may not be intended by those making the rules. Conversion takes place when actors repurpose an existing institution for goals beyond its original intent. The authors argue that these forms of change are typically exploited by organized interest groups and politicians, rather than by voters. They also point out that greater precision in institutional rules makes them more susceptible to drift, while greater ambiguity makes institutions more vulnerable to conversion. They note as well the tradeoffs involved in automatic procedures, which tie the hands of future rule-makers but can be difficult to enact and can have unintended results, and delegation, which adds flexibility but can create principal-agent problems.

The authors note, but could have further emphasized, that American politics are unusually subject to drift and conversion due to strong partisan divisions and institutional procedures that require high consensus for change. They point out that each US state gets two senators even though the most populous state has 65 times as many people as the least populous (p. 187), but they could have added that this ratio was only 11 to 1 when the constitution was written. This has undemocratic implications for policies on which urban and rural preferences diverge, including farm subsidies, energy policies, and gun control.

A third theory-focused chapter by Giovanni Cappocia focuses on critical junctures and path dependence. Cappocia outlines the standard version of path dependency—critical junctures are periods of contingency in which actors make choices that create long-term institutional trajectories—but he adds several useful qualifications. First, he brings agency back in by noting the role of policy entrepreneurs in manipulating the normative framing of proposed changes to shape other actors’ preferences and assemble winning political coalitions at critical junctures. Second, he acknowledges that the initial preferences of the most powerful actors do not necessarily determine institutional outcomes. Third, he notes the methodological importance of studying unrealized critical junctures where institutional change could have taken place but did not. Cappocia acknowledges in a footnote that his chapter focuses only on path dependencies involving increasing returns (p. 148).

A section on methodology includes a chapter by Mahoney and Falleti that focuses on the comparative sequential method, which combines cross-case comparisons and within-case analysis (process tracing). This chapter focuses not only on self-reinforcing sequences but also on the kind of reactive sequences that the rest of the book sets aside. It distinguishes between causal sequences (A leads to B leads to C) and strictly temporal sequences, in which events themselves are not causally connected but the order, duration, pace, or timing of events causally affects the outcome. The chapter nicely summarizes these distinctions in tables on the types of sequential arguments in CHA and the types of processes in CHA, and it discusses how to use inductive/theory generating and deductive/theory testing process tracing as well as case comparisons to develop and assess CHA arguments.

A second methodology chapter by Evan Lieberman discusses how to combine quantitative, experimental, quasi-experimental, and case studies in CHA analyses. Lieberman outlines how to use quantitative analysis to assess alternative explanations, make statements about populations, and help guide case selection, and how to combine this with case studies to improve concepts and measures and assess whether hypothesized mechanisms were in fact operative in specific cases. Lieberman also analyzes the contributions that matching, experiments, and natural experiments can make to CHA. One limitation of this otherwise very useful analysis is that Lieberman emphasizes starting with large-N analysis and then moving to case studies, whereas at times it may be useful to start with case studies to refine concepts and measures before coding many cases for statistical analysis. The chapter could also benefit from Thad Dunning’s analysis of how case studies can help assess the degree to which natural experiments meet the assumption that assignment to treatment versus control groups is “as-if random” (“Improving Process Tracing: The Case of Multi-Method Research,” in Andrew Bennett and Jeffrey T. Checkel, eds, Process Tracing: From Metaphor to Analytic Tool, 2015).

An epilogue by Wolfgang Streeck usefully situates CHA in comparison to earlier traditions in historiography and their iconic practitioners. Streeck maintains that in contrast to Thucydides, Machiavelli, Ibn Khaldun, and Marx, respectively, CHA brings to the fore structural constraints rather than individual actions, engages in systematic comparison for causal analysis rather than only seeking particular historical lessons for policy-makers, abjures universal laws of societal development, and views history as contingent rather than teleological. Streeck concludes that CHA thus constitutes a distinctive approach to history that is likely to continue to contribute to our understanding of politics. This is a fitting conclusion to an excellent volume that should find its way into many graduate courses on comparative politics and research methods.