Defining civilian control over the military is relatively straightforward: do civilian leaders make decisions about the military, and does the military obey? But measuring civilian control is another matter. One can of course observe whether the military refused to obey an order. But when the military obeys, does it truly reflect civilian control, or simply reflect a concordance in preference over that order, or worse yet—reflect that the civilian had already acquiesced to the military’s preference? The key to civilian control is whether the military obeys orders that it does not otherwise want to obey, a challenging judgment call to make.

That difficulty in measuring civilian control is one reason why our field has for decades lacked a cross-national quantitative measure of it (for a recent exception see Michael R. Kenwick, “Self-Reinforcing Civilian Control: A Measurement-Based Analysis of Civil-Military Relations.” International Studies Quarterly 64:1, 71–84, 2020), and scholars have instead focused on more easily observable outcomes like coups and defections (for more on this observation, see Risa Brooks, “Integrating the Civil–Military Relations Subfield.” Annual Review of Political Science 22, 379–398, 2019). The only way to truly measure civilian control would be to do deep, qualitative research about every case—and to do so in a systematic fashion across all countries.

Kuehn and Croissant have admirably set out to do just that and have made an enormous contribution in the process. Their new book, Routes to Reform, introduces one of the first-ever measures of civilian control across 66 democracies in the third wave (1974–2010). They then use it to explore why some new democracies have been better than others at establishing civilian control and the consequences for both democratic survival and quality.

Kuehn and Croissant also offer a disaggregated measure of civilian control across five domains: (1) Elite Recruitment (measuring whether military officers control or assume political office), (2) Public Policy (measuring military influence over non-security/defense-related policies), (3) Internal Security (measuring the military’s involvement and control over internal security operations), (4) National Defense (whether civilians devise and decide defense policy), and (5) Military Organization (whether civilians decide military doctrine, education, procurement, promotions, and so on). Intuitively, they show that while third-wave democracies have been quite good at establishing civilian control on the first two dimensions, the latter three—that strike at the heart of the military profession—have proceeded slower, if at all.

After introducing their measure of civilian control, the bulk of the book then seeks to theorize and test both the causes and consequences of civilian control. First, Kuehn and Croissant argue that the key to establishing civilian control is whether new democracies pursue what they call “robust control strategies,” which involve punishing and sanctioning violators; counterbalancing or playing off internal rivalries; and developing monitoring and oversight mechanisms over the military. By contrast, other new democracies pursue only “weak control strategies,” namely, socialization into democratic values; screening, selecting and promoting officers who align with civilian preferences; and appeasing and acquiescing to military demands.

Of course, a new democracy’s ability to pursue robust strategies is a function of a bargaining struggle with the military—if it pushes farther and faster than the conditions allow, it might elicit a coup. Kuehn and Croissant thus outline 11 factors that might shape this bargaining struggle, including factors related to initial conditions (praetorian legacies, military control over the transition), factors related to the civilian’s bargaining power (institutionalization of democracy, civil society, international support), and factors related to the military’s bargaining power (internal and external security threats, economic development, socioeconomic inequality, elite conflicts, and mass discontent).

Notably, in their quantitative analysis (Chapter 3), only two of these eleven factors reach statistical significance: the strength of civil society and economic development. Those, they conclude, are the most important in establishing civilian control. However, the remaining factors do play a role in the case studies Kuehn and Croissant subsequently conduct in Chapter 4 (Chile, Spain, South Korea, Madagascar, Philippines, and Russia). Similarly, in the quantitative analysis, Kuehn and Croissant do not actually measure the robust v. weak control strategies, nor show that those are indeed the mechanism by which economic development, for instance, shapes civilian control. But here again, their case studies prove useful, showing which strategy each democracy chose and why.

Finally, Chapters 5 and 6 move towards testing the effects of civilian control. Kuehn and Croissant show that civilian control exhibits a significant correlation not only with democratic survival but also with the quality of democracy, particularly for civil and political rights. Although the mechanisms and causality are difficult to establish (for instance, it could be that democracies with more political crises can establish neither civilian control nor high-quality institutions), Kuehn and Croissant are once again able to leverage case studies to illustrate their theory. In Chapter 6, they provide brief narratives of all 18 cases of democratic breakdown in their dataset, as well as a paired comparison of Brazil and Taiwan to show the effects on democratic quality.

Overall, Kuehn and Croissant advance a relatively optimistic conclusion: that past is not prologue. Democracies that inherit powerful, praetorian militaries are not necessarily doomed to failure. If they develop strong civil societies, generate economic growth, and happen to be blessed with prudent and patient civilian leaders, they can pursue careful yet robust control strategies that help them develop high levels of civilian control.

While true, it is important to note that praetorian legacies might have exhibited a stronger effect in their analysis had they pursued a more universal set of cases. Kuehn and Croissant exclude short-lived democracies that only survived a year or two, nor do they consider countries Polity codes as in a transition. This selection makes sense for the primary focus of their analysis, which is on the longer-lasting democracies. But, given that coups tend to occur early in a transition, many of the praetorian cases might have already collapsed before entering into their dataset.

A second major contribution Kuehn and Croissant make is to highlight that militaries matter not only in coups but also in other forms of democratic breakdown, including executive aggrandizement. They also helpfully distinguish whether militaries play a partner v. bystander role in civilian aggrandizement (see especially pp. 176–185). Moreover, they uncover two intriguing correlations with whether militaries play a partnering rather than bystanding role in breakdown. First, partners tend to have higher levels of civilian control than bystanders (p. 186), potentially in line with an argument I recently made in my own book about how civilian control can also have its dark sides, with such militaries being more likely to obey orders to, for instance, shut down a parliament (Grewal, Sharan. Soldiers of Democracy? Military Legacies and the Arab Spring, 2023). Second, they find that militaries are more likely to play a partnering role if they are involved in and have some control over, internal security roles (p. 188), harkening back to O’Donnell and Schmitter’s observation that these roles are the most dangerous in politicizing the military (O’Donnell, Guillermo and Philippe C. Schmitter. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies, 1986, p. 31).

A final lesson Kuehn and Croissant draw is that gradual, prudent, trial-and-error approaches are the key to establishing civilian control (pp. 212–214). It may have been useful to also show this quantitatively—to see if attempts to establish robust control strategies when the conditions were not propitious indeed sparked democratic breakdown (and more generally, it might have been useful to show that bargaining powers themselves shaped democratic survival). Indeed, the case studies in some ways qualify this conclusion that new democracies should move slowly. South Korea’s Kim Young-Sam, the authors note, made sweeping purges immediately upon taking office (p. 99), and Spain and Chile seemed to make similarly dramatic moves as soon as the opportunity arose (i.e., post-failed coup and post-Pinochet arrest, respectively). Perhaps the key is not to move slowly per se, but to be ready to pounce when the opportunity arises, whether that is early or late in a transition.

Overall, Kuehn and Croissant have offered the field a remarkable new dataset on civilian control, and one that should become a staple in quantitative analyses of democratization and democratic breakdown. Substantively, the book also provides crucial insight into how to establish civilian control, deriving lessons learned from illuminating case studies. This is an essential read for political scientists and policymakers alike.