Perhaps it is too bold a statement to claim that studying violence and leadership always produces compelling scholarship, but it is certainly true for In Time of War by Adam Berinsky and Democracy at Risk by Jennifer Merolla and Elizabeth Zechmeister. Both books strike me as must-reads for scholars of political behavior, as well as for graduate students preparing for comprehensive exams. In addition, both are well suited to either advanced undergraduate lecture courses or graduate seminars. (I assigned the first in an undergraduate course on war and public opinion and will now add the second, too.)

These books share a common theme—citizen response to unusual and violent times. Berinsky examines citizen attitudes and behavior during times of war, while Merolla and Zechmeister examine how citizens cope with terrorism and the threat of terrorism. The two volumes are complementary in a number of important ways. Berinsky examines how citizens respond when the United States engages in violent conflict with other states; Merolla and Zechmeister examine citizen response to the threat of violence fomented by nonstate actors. Berinsky examines how citizens are influenced by political elites; Merolla and Zechmeister examine how conditions of threat lead citizens to view leaders differently, including projecting charisma and other leadership qualities onto political elites (or at least a single political elite). Berinsky resurrects a veritable treasure trove of public opinion survey data from the 1930s and 1940s; Merolla and Zechmeister design a series of clever experiments.

Despite the common theme and complementarities, there is also an important difference. Berinsky argues that there is nothing terribly special about times of war when it comes to public opinion. While war may be a transformative event in many regards, it does not change how citizens think: “Public opinion in times of war is properly viewed as a continuation of the same processes that shape public opinion during times of peace” (p. 2).

Merolla and Zechmeister, on the other hand, argue that unusual times can powerfully affect citizen behavior and thought processes: “Our principal message and findings can be simply stated: the specter of terrorist threat results in attitudinal, evaluative, and behavioral shifts, some of which can potentially endanger democracy” (p. 2).

The two books together are extremely helpful in painting a picture of how citizens and polities respond to uncertain and/or threatening times. Sadly, that picture is not pretty.

Berinsky's core argument is that the same factors that drive support or opposition to domestic policies also drive support or opposition to foreign policies. Thus, partisanship and social group attachment are key causal factors in understanding citizen support for war. Citizens use partisanship and group attachments as cues or heuristics for foreign policy, just as they would for domestic policy. As part of this argument, Berinsky believes that the evidence showing that citizens respond to the “events” of war is vastly overstated.

While he does not make it a central tenet of his book, Berinsky provocatively argues that not only is citizen opinion driven by the stand of partisan elites; elites also decide where to stand on matters of war and peace largely for domestic political reasons (pp. 85–86). Taken to the extreme, this account is normatively horrifying. U.S. decisions to engage in war and peace have little to do with the substance of international relations or protecting national interests; public support for these decisions are made simply by following who is advocating what policy, and the specifics of a (possible) war are of no consequence.

I find this vision to be too grim. If Berinsky's main criticism of the events literature is that it fails to appropriately take into account the ability of political elites to frame or even ignore what is happening on the ground, the flip side of that argument applies to In Time of War. Simply put, elites do not have an unfettered ability to manipulate. A large body of work shows that citizens are sensitive to mission goals. While that does not limit elites per se, it does mean that what elites say matters in addition to who is saying it. Some of Jamie Druckman's recent work showing limits on the ability to frame also seems relevant here. The situation on the ground likely constrains what elites are able to credibly say or how they can frame military conflict. Berinsky recognizes that conditions on the ground can matter in certain circumstances, such as influencing election outcomes: “Success [on the battlefield]—or more precisely, the absence of success—can matter in elections” (p. 191).

For me, the strongest parts of the book are those that utilize the survey data from the 1930s and 1940s. Much has been written on public opinion regarding the Vietnam and Iraq Wars; novel contributions are necessarily smaller. Compelling public opinion analyses for citizen attitudes towards World War II are more or less nonexistent. It is not often that we get brand new data from seventy years ago. I am especially impressed with the focus of the early chapters devoted to debunking myths related to public opinion toward World War II (such as the attack on Pearl Harbor suddenly and dramatically moving public sentiment from isolationism to support for war).

Turning to the other book in this review, Merolla and Zechmeister have written what may become a definitive volume on citizen response to terrorism. They make three central theoretical contributions: 1) Terrorist threat activates those with greater authoritarian predispositions to support more authoritarian policies; 2) terrorist threat causes citizens to look for charismatic leaders, but also causes citizens to project qualities of leadership onto candidates (and then weight “leadership” more heavily when making a vote choice); and 3) terrorist threats generate support for more activist foreign policies.

All three claims are well supported by the experimental evidence. In their analysis of the experimental data, the main variable of interest is assignment to a “terrorist threat” treatment condition that emphasizes the possibility of an imminent attack. Depending on the experiment, this treatment is compared to a rival “good times” treatment emphasizing economic prosperity (some experiments also included an “economic threat” treatment, a pure control condition, or both). The authors conduct a series of similar experiments across a number of contexts, all with fundamentally similar results. These differences include student and nonstudent samples, audiovisual treatments as well as text-only treatments, and studies done in both the United States and Mexico.

The most interesting and most compelling chapters are those that focus on leadership qualities. What might be of most general interest is the claim that charisma is less a trait that one possesses than something that is highly malleable by context. Threat causes citizens to see (certain) leaders as more charismatic. Reading these books side by side makes me ponder how taking into account Merolla and Zechmeister's findings that perceptions of leaders change with circumstances affects the cues that elites provide to voters. Neither book directly provides evidence either way when their theoretical narratives overlap like this.

As with Berinsky, the consequences of Merolla and Zechmeister's findings are troubling. Taken to the extreme, democracy and civil liberties are luxuries that we enjoy only when there is no threat. Once the threat is severe enough, we cannot expect these institutions to endure. Rather, we can expect to look to a leader to save us. The argument is eerily reminiscent of Hobbes' Leviathan, where subjects willingly submit to the absolute power of a sovereign as their best means of protection against a cruel and unforgiving world.

I do have a pair of minor quibbles, neither of which fundamentally alter my positive opinion of Democracy at Risk. First of all, I would have liked to see a “good times” condition that focused on directly decreasing perceptions of terrorist threat, not just emphasizing positive economic conditions. Bringing additional considerations, like the economy, into the mix slightly blunts the power of otherwise excellent experiments. Second, when the experimental conditions are included in interactions terms, I would also like to see plots of the marginal effects of the treatments (and confidence intervals) across the values of whatever it is being interacted with.

In sum, these are a pair of excellent books that should be on the shelves of political behavior scholars (especially where behavior somehow intersects with international relations) as well as on both graduate and undergraduate syllabi. That being said, the conclusions of both are extremely troubling.