As the title of his new book implies, Daniel DiSalvo sees party factions as “engines of change,” the prime movers in American politics. If we want to understand nominations, key policy decisions, and the growth of the state, DiSalvo argues, our focus should not be on conflict between the parties or among significant individuals but somewhere in between.

A difficulty in studying factions is the absence of a clear roster. Factions do not show up in election returns or official legislative documents. The first contribution of DiSalvo's study is thus his careful compilation of a list of US party factions since the Civil War. The author identifies factions on the basis of four criteria: ideological consistency, organizational capacity, temporal durability, and the ability to attempt to shift the party along the right–left spectrum. Compared to other ways that the term “faction” has been used in political science, these criteria may seem somewhat restrictive. Factions in Japan's Liberal Democratic Party, for example, are famously nonideological, as were the factions identified by V. O. Key in Southern Politics (1984). But requiring ideological consistency keeps the focus on the most significant factions and still produces a set of factions large and diverse enough to characterize the various ways they have impacted national-level politics over a century and a half. Moreover, by focusing on groups with an identifiable ideology linked to the party's right position, DiSalvo distinguishes factions from the more numerous groups associated with narrow policy demands.

The author combed newspapers, party documents, and historical scholarship in order to identify 12 factions that meet his criteria, five in the Democratic Party (Populists, Southern Democrats, Liberal-Labor, New Politics Democrats, New Democrats) and seven among the Republican (Stalwarts, Mugwumps, Half-Breeds, Old Guard, Progressives, Liberal Republicans, New Right). This systematically compiled list is a resource that other scholars will find useful. These factions are diverse along many dimensions: size, longevity, and goals. Just over half are classified as change seekers, a quarter as preservationists, two as a mix.

Studying how these factions have behaved in various domains, DiSalvo paints a vivid picture that shows how they have shaped American politics. He writes (p. 30) that

For example, factions have been “conveyor belts of ideas,” reconfiguring party ideology and policy agendas. In some cases, this has been relatively straightforward, by way of illustration, as the ideology of the New Right became the dominant ideology of the Republican Party as a whole. The Progressive ideology, on the other hand, followed a more convoluted path, DiSalvo shows, as it moved from a strong and vibrant Republican faction to a splinter party, finally seeing its greatest impact under Democratic President Woodrow Wilson.

Factions are often active in presidential nominations. The book's analysis spans 33 presidential elections, thus 66 major party nominations. Twenty-three of these nominations were won by candidates clearly associated with factions. Some factional candidates, such as Ronald Reagan and Bill Clinton, did well in general elections; others (George McGovern, Barry Goldwater, William Jennings Bryan) did not. Factions also play an important gatekeeping role in nomination contests, vetoing candidates who are ideologically unacceptable or too closely aligned with opposing factions. DiSalvo notes the value placed by some factions on nominating candidates whom they can trust. He quotes as an operative saying, “the one thing the AFL-CIO can't forgive McGovern for is the one thing he can't do anything about: if he's nominated, he won't owe labor anything” (p. 82). Indeed, the worry that a successful candidate might betray the factions that supported him was not misplaced, as evidenced, for example, by Rutherford Hayes. The Stalwart Republicans had accepted Hayes as a compromise candidate, not aligned with any of the party's three factions, only to see his administration enact the civil service reforms they had most vehemently opposed.

Disappointed factions sometimes split from their parties, either to run third-party candidates (as the Progressive Republicans did in 1912 and the Southern Democrats in 1948 and 1968) or to vote for the candidate of the opposite party (as did the New York Mugwumps in 1884). From a party-centered point of view, running a third-party candidate can seem irrational. But, as DiSalvo shows, from a faction-centered point of view, it can make sense. Yes, by running Theodore Roosevelt as a third-party candidate, the Progressive Republicans contributed massively to the victory of Democrat Wilson. But, as noted, Wilson enacted key aspects of the Progressive agenda in a way that William Howard Taft almost certainly would not have. Other factional defections (Mugwumps in 1884, Southern Democrats in 1968) could also be seen as strategic voting from the point of view of policy preferences. Even Strom Thurmond's 1948 third-party candidacy can be seen as strategic: If Harry Truman had lost (certainly a plausible outcome ex ante), one suspects that the southern faction would have emerged with more leverage against the rest of the Democratic Party.

In the course of pursuing their ideological goals, factions have shaped the internal organization of Congress, seeking strategic advantage via greater centralization or decentralization of power as circumstances dictated. Factions have structured the challenges and opportunities that presidents face in the pursuit of policy agendas. Factional conflict has propelled the major policy initiatives that have defined the development of the American state: Reconstruction, civil service, the major waves of economic regulation in the early twentieth century, and the new social regulation of the post-Vietnam years.

Current theories of political parties view them as forces that stabilize the potentially chaotic process of coalition formation. Scholars may disagree about the nature of party coalitions, but there is general agreement about the stability they promote. DiSalvo's study of factions shows, however, that far greater coalitional stability occurs at the factional level. From a social choice perspective, then, factions are more stable than parties. Members of a faction support each other more consistently than they support copartisans outside the faction. This statement verges on tautology: What could we mean by “faction” if not that? What the author shows us, however, is how important factions are. Jockeying and shuffling among factions is how competition among ideas and interests in the broader society reaches the institutions of government

From a temporal perspective, however, factions are less stable than parties. Factions have shorter lifespans: DiSalvo estimates the durations of those he studies as ranging from 18 (Liberal-Labor Democrats) to 42 (Southern Democrats) years. The Democratic and Republican parties, in contrast, have endured for the century and a half spanned by his study. The lifespan of his factions is, however, comparable to the lifespan of parties in many countries that use proportional electoral systems. One might easily conjecture that a coalition of interests that remains a faction in the United States would be its own party in other countries. But parties (even small ones) in proportional systems are often themselves factionalized (e.g., the small German Green Party was divided into “Realo” and “Fundi” factions in the 1980s and 1990s.)

The ways in which party factions vary across countries and institutions is, of course, beyond the scope of Engines of Change, but it exemplifies an important feature. The book is a conversation starter. The focus on a single country and a limited (though far from short) time period allows a level of detail that would not be possible with a broader scope. The detail and nuance in these accounts of factional impact draws our attention to broader questions. When do parties nominate factional candidates? When do counterfactions organize? When does an ideological movement beget a faction? Are policy proposals more likely to succeed if they are promoted by a faction? To answer these questions, future studies may augment DiSalvo's systematically compiled lists of factions and factional candidates with complementary events and observations: unified parties, nonfactional candidates, ideologies that had impact in the absence of an associated party faction. The book gives these future projects a basis to build on and a reason to incorporate factions into our understanding of party politics in a systematic fashion.