In this essay we review the recent history of “social sorting” in American politics. We describe how partisan identities have grown increasingly aligned with other social identities such as race, religion, ideological identity, region, and culture. We connect this phenomenon with research from social psychology and comparative politics on the psychological and sociological effects of this type of identity sorting. In particular, well-aligned identities increase intolerance of political and social outgroups, and societies politically divided along ethnic and/or religious lines are at greater risk of descending into civil war. In fact, early American political scientists suggested that the unique stability of American democracy lay in the cross-cutting (unaligned) nature of American political and social identities. We build on these theoretical connections with new findings from our own research on the depths of partisan animosity in the U.S. today. We find that Democrats and Republicans vilify their partisan opponents in the strongest terms, they feel less negatively when those opponents physically suffer, and a small minority advocate outright violence against them. We situate those modern hostilities in a broader American historical context. Finally, we discuss the normative implications of these findings. The obvious negative implications extend to the risk of widespread civil unrest and escalating political violence. One positive implication emerges, however. Under the current partisan alignment of racial and religious identities, the overwhelming divide between the parties is between traditionally high-status (e.g. white, Christian) and traditionally marginalized (e.g. non-white, non-Christian) groups. This provides unprecedented power to previously marginalized groups – an entire political party pressing for their advancement. The social unrest that we currently see can therefore either be seen as the path toward violent division, or as a step toward an inherently disruptive movement toward social justice.

Chapter 6 analyzes the impact of casualties on elections during the war. Statewide elections were staggered throughout the calendar in ways that make those election returns function like a tracking poll of partisan support. Contrary to modern war studies, I find Republican vote shares did not deteriorate in response to national casualties. I validate my analytical approach with evidence of highly nationalized and synchronized elections linking party votes for House, governor, and president both before and during the war. However, local deaths moved voters on the margins, with disparate impacts based on the balance of local partisanship. Local deaths depressed Republican vote share in prewar Democratic bastions but not in Republican strongholds. Recent deaths were especially potent. It suggests that the combination of partisan identities, local partisan leaders, and local social environments shaped local interpretations of the war – including its unprecedented human costs. Local casualties did change the electoral behavior of some voters, while most stood pat. Partisan mechanisms guided who responded and how.

Leaders play a key role in whether their followers accept election losses or turn to violence instead. Here I illustrate the importance of public concession by party leaders after election losses, comparing 2008 and 1860.

I conclude the book by reviewing the main findings and their broader implications for mass partisanship and violence beyond the Civil War era. The essential ingredients for partisan conflict have been present throughout most of American history: strong political identities placed in competition. Yet, the U.S. has been mostly free from partisan violence over the past half-century despite partisan animosity, electoral discontent, and even some public support for violence. What changed in the Civil War era was how partisanship aligned with other important social identities, how party leaders fueled the crisis, and then how parties explicitly organized legitimated violence via party control of government. The country is in an uncomfortably similar position today – racial-religious-partisan alignments, political demonization, rhetoric rejecting fair elections, and even language encouraging violence from prominent leaders, including the president. I draw out those implications here.

Chapter 8 investigates postwar partisanship, first in enduring postwar election effects from local casualties, then in war memorialization, and finally in the partisan dynamics of Union veterans’ organizations. I describe the political aftermath of the war and test whether wartime voting patterns – including casualty effects – persisted after the fighting and dying ended. I find the massive scale of wartime death continued to shape local political voting patterns in postwar presidential elections for decades, a first in casualty-opinion studies. The chapter also shows how partisanship predicted which states were first to commemorate their dead on Decoration Day, to document their state’s war service in bureaucratic records, and to honor their veterans and the dead with local monuments. Chapter 8 also finds more chapters of the Grand Army of the Republic (GAR) veteran’s group in prewar Republican places, along with evidence of the GAR’s indirect impact on General Ulysses S. Grant’s 1868 election, when veterans organized actively to boost Republicans.

Chapter 2 begins by describing how prewar partisan politics and violence led to mass partisan warfare. Next, I detail the theoretical foundations of mass partisanship under duress, drawing on research from political science and psychology. Lastly, I show the persistence of partisan voting patterns across party systems in the North – surprisingly stable electoral coalitions despite changes in name and issue emphasis. That continuity provides a solid basis for predicting strong partisanship in individual identities, leaders, and communities during the war.

Chapter 1 introduces the historical context and key partisan questions, then it outlines the book’s methods and contributions in light of diverse social science and history literatures. Finally, I outline the plan of the book and summarize the approach and results in each chapter.