Numerous studies attempt to explain institutional change in the U.S. House of Representatives. Much of this literature adopts a House-centric approach whereby the chamber’s rules are a product of the preferences of its members and the strength of party leaders, with the latter often endogenous to the former. In Separation of Powers and Legislative Organization, Gisela Sin argues that in order to understand when and why the House changes its rules, we must consider the preferences of outside actors: the president and the Senate.

In Sin’s framework, rules should be viewed “in terms of reallocating power between the groups within the majority party” (p. 12). The author models the decision to change the rules of the House as a bargaining game among various House factions, the Senate, and the president. She develops a formal model in which House factions choose to allocate power with an eye toward how these rules will affect the sorts of policies that can be enacted, given the preferences of the Senate and the president. As a result, some institutional changes in the House come about due to the election of a new president or a party shift in the U.S. Senate. Sin’s Constitutional Theory of Legislative Organization, then, makes the novel claim that the way in which power is allocated in one chamber and the timing of institutional changes can be a function of the preferences of external actors.

Five important groups populate Sin’s theory: two majority-party factions (the Speaker’s faction and the non-Speaker faction), the minority party, the Senate, and the president. In Chapter 2, the author describes the various factions that have existed over the period of 1879–2013. Importantly for her theory, she identifies which majority-party faction was closest to the Speaker during each Congress, as well as which House faction was closest to the president.

In Sin’s account, House members are strategic actors who look to the preferences of external actors when deciding how to bargain and allocate power among themselves, and their utility is a function of policy outcomes. Once power has been allocated, members engage in a legislative game in which each actor knows the preferences of other actors. In this complete-information game, Sin uses backward induction (starting from the legislative stage) to compute equilibrium predictions pertaining to the power-sharing stage.

The first empirical implication of Sin’s theory is that rules may change when the preferences of the Senate and/or those of the president change. That is because a change in these external actors’ ideal points results in a shift of the policies that can beat the status quo. House factions will then adopt rules that will result in the best policy outcomes given the set of feasible enactments. To test this claim, the author first identifies every major rule change between 1879 and 2013. She then shows that, consistent with her theory, rules tend to be modified when the ideal points of the key actors change. For example, when partisan control of the Senate or the presidency changes while the House factions remain constant, House rules are changed 83% of the time, which is more than twice the rate when there is no change in the preferences of any actor. This demonstrates that House members may have an incentive to change the rules of the chamber even when their preferences remain unchanged. This is a major theoretical departure from existing work, and it underscores the importance of Sin’s book.

In Chapter 5, the author tests another theoretical implication regarding the direction of institutional changes. She finds that power tends to be centralized in the hands of the Speaker’s faction when the other (non-Speaker) majority-party faction is closer in terms of policy to the Senate or the president than to the Speaker’s faction.

I find this book to be an exemplar of scholarship that is theoretically driven, empirically rigorous, and historically sound. Not only does Sin’s account explain much of the institutional variation observed over a long time period, but the author also revisits well-known historical episodes of congressional upheaval and institutional change. For instance, in Chapter 6, Sin reexamines the events that led to a coalition of Democrats and Progressive Republicans who stripped power away from Speaker Joe Cannon in 1910. She shows that the election of William Howard Taft in 1908 played a very important role, as Progressives no longer had an ally in the White House. Prior to the 61st Congress, Progressives could count on President Theodore Roosevelt to block legislation that went too far in the conservative direction. Here is a perfect example of her theory at work: forward-looking House factions reallocating power following a change in the policy preferences of an external actor.

What I found particularly interesting is that the book raises additional questions, which I take to be the hallmark of good, thought-provoking scholarship. For instance, Chapter 6 shows that the election of a president who shares the policy preferences of the Speaker may lead to a reallocation of power away from the Speaker’s faction. This means that it is possible that the policy goals of congressional leaders are in conflict with other goals, such as holding power. This is not a concern for Sin’s theory because her actors are motivated solely by policy. But it does raise the question of the extent to which congressional leaders are willing to sacrifice other goals in the pursuit of policy.

One aspect that is missing from the story is the role of elections and how they may affect the allocation of power. It seems that even if members care only about policy, Sin’s own accounts suggest that something else may also motivate them. For instance, the model depicted on page 81 shows a power-sharing arrangement between both majority-party factions in which power is decentralized away from the Speaker. Yet both the non-Speaker faction and the minority party (which together comprise a majority of the chamber) would be better-off if they agreed to form a coalition. Why does the non-Speaker faction not follow through with such a cross-party agreement, which would make it better off in terms of policy? Similarly, the centralizing of power depicted on page 80 suggests that the non-Speaker faction will agree to centralize power in the hands of the Speaker, even though it is indifferent to three types of arrangement. Why would the minority party, which is much worse-off under this centralized agreement, not offer inducements so that the non-Speaker faction will agree to a cross-party coalition? Perhaps the answer is that there is a cost for a majority-party faction to form a coalition with the minority party, which suggests that there may be a role for elections after all (since elections may constrain the extent to which legislators may wish to dilute the party brand, for instance). Or perhaps parties constrain member actions in ways that go beyond policy.

None of this should detract from this impressive piece of social science. The book should be of interest to scholars of legislatures and political development. It furthers our understanding of legislative institutions and explains some institutional developments in a way that extant accounts could not. It should be required reading for anyone who wishes to understand when, why, and at whose expense politicians will reallocate power.