Ellen M. Immergut, “The theoretical core of the new institutionalism”, Politics and Society 26 (1998), pp. 5-34

Summary: Outlines the institutionalist tradition, tracing its roots back to behaviouralism. Institutionalism is divided into three variants, as per Hall and Taylor (1996). A common theme in all of the variants is then presented.
Important Insight: The three variants of new institutionalism share a common goal. All are concerned with the difficulties of ascertaining what human actors want when the preferences expressed in politics are so radically affected by the institutional contexts in which these preferences are voiced. Rather than tackling this question by probing individual psychology, these scholars have turned to analyzing the effects of rules and procedures for aggregating individual wishes into collective decisions.
—-
Notes
-key point: the three variants of new institutionalism share a common goal. All are concerned with the difficulties of ascertaining what human actors want when the preferences expressed in politics are so radically affected by the institutional contexts in which these preferences are voiced. Rather than tackling this question by probing individual psychology, these scholars have turned to analyzing the effects of rules and procedures for aggregating individual wishes into collective decisions.
-the focus on observable behaviour under Behaviouralism is the point of departure for the new institutionalism. The new institutionalists vehemently reject observed behaviour as the basic datum of political analysis; they do not believe that behaviour is a sufficient basis for explaining ‘all of the phenomena of government’. For behaviour occurs in the context of institutions and can only be so understood.
-there are three aspects to the institutionalist critique of behaviour.
-The first questions the assumption that political behaviour reveals preferences (true interests cannot be determined by action, since there may be a distinction b/w ‘expressed’ and ‘real’ preferences). Institutionalists theory aims to expose and analyze the discrepancy b/w ‘potential’ interests and those that come to be expressed in political behaviour.
-Second, the institutionalist approach views the summation of preferences – or, for that matter, the aggregation of individual behaviours into collective phenomena – as exceedingly problematic. Human interests are so complex that to speak of summing or aggregating them is merely applying a metaphor to a complicated process. Mechanisms for aggregating interests do not sum but in fact reshape interests – by developing new ideas through discussions and getting some persons to redefine their preferences, by selecting out some interests at the expense of others, or by reducing a multifaceted set of issues to two alternatives that can be voted on.
-The third institutionalist challenge is normative. If the institutionalists are correct, much or all of political behaviour and collective decision making is an artifact of the procedures used to make decisions. If political processes are seen to be this decisive, the analyst’s evaluation of politics will change. Interests will no longer be regarded as subjective assessments of individuals; collective decisions will no longer be regarded as subjective assessments of individual wishes.
-The recognition of bias in institutions, however, burdens the institutionalist with two responsibilities. Institutionalists should discuss the direction and implication of this bias, and they should suggest ways to improve the justness of institutional outcomes. Yet for reasons related to institutionalist assumptions themselves, these challenges are extremely difficult to meet.
-by rejecting both a behaviouralist and a social determinist/Marxist understanding of preferences (with the former accepting (a posteriori) the expression of interests at face value and the latter adopting an objective standards of justice (stemming from class, gender, or social position a priori)), institutionalists are forced to ‘square the circle’ b/w a priori and a posteriori standards by recommending formal procedures that can be used to define substantive justice.
-Analyses of existing procedures and their distortions provide guidelines for these institutional recommendations. But institutionalism cannot provide a positive theory of standards that can be used to evaluate political choices and outcome.
-how should we understand and interpret political choices according to institutionalists?
-rational choice institutionalism: institutions allow political choices to be made b/c they do not allow every conceivable political choice to be considered. Moreover, b/c political actors are aware of the effects of these rules, they will attempt to cast votes or to manipulate the rules in such a way as to achieve their most-preferred outcome. Consequently, voting expresses not the true preferences of voters but an indeterminate amalgam of honest and strategic voting.
-individual utilities serve as the standard for judging political institutions and outcomes.
-organization theory (sociological institutionalism): political decisions or any other decisions cannot be understood as macro-aggregations of individual preferences but instead result from cognitive and organizational procedures that produce decisions despite uncertainty.
-rejects utilitarian assumptions about the satisfaction of individual preferences and interests through collective decisions. Lacks a standard for judging political institutions.
-historical institutionalism: by tracing changing definitions of interests through time and across cultures, the impact of institutions on the construction of interests can be studied without imposing arbitrary, ‘objective’ definitions of interests. That is, the discrepancy b/w ‘potential’ and ‘expressed’ preferences can be addressed without inventing a theory of the actors’ ‘true’ interests.
-as a corrective to structuralism, methodological individualism is used even by scholars who analyze collective actors; human agency is better integrated with structural factors; and the role of ideas has been given greater weight. As a reaction to a greater interest in interpretation, three themes (which can be traced back to Weber) have become even more central in this work: (1) these scholars are interested in ‘alternative’ rationalities; (2) causality is viewed as contextual, with complex configurations of factors being causally significant; (3)  the group emphasizes the contingencies of history (chance plays a large role, and can have lasting effects)
-the political construction of interests: institutions (be they formal rules of political arenas, channels of communication, language codes, or the logics of strategic situations) act as filters that selectively favour particular interpretations either of the goals toward which political actors strive or of the best means to achieve these ends.
-three issues of historical institutionalism are particularly troubling to me: (1) the problem of falsifiability: without a sufficiently broad comparative perspective, historical institutionalists risk overstating the uniqueness of their case, and it is difficult to see how such historical narratives can ever be proved wrong; (2) historical institutionalists profit somewhat unfairly from the positive models that they criticize, since their studies are often too inextricably tied up to their original context to be subject to the generalizabilty that makes the theories they criticize criticisable; (3) in eschewing systematization, the historical institutionalists undercut the cumulative impact of their work.
-we should seek to develop the historical institutionalist analysis of power, and to do this we will have to do more than attack essentialist and determinist conceptions. It is not at all clear how we can develop a nondeterminist concept of power, but I think it is worth the effort.