How do states die? That is an important question to tackle and has been frontline news again with Russia’s invasion of Ukraine. Systematic empirical analyses of state death took off with Tanisha Fazal’s pathbreaking work (State Death: The Politics and Geography of Conquest, Occupation, and Annexation, 2007; “State Death in the International System,” International Organization, 58(2), 2004). Fazal defined state death and analyzed its causes quantitatively, distinguishing four types of sovereign state death: conquest, occupation, voluntary unification, or dissolution. Many other studies have followed in her footsteps and employed Fazal’s definition of state death.

Doug Lemke’s new book contributes to this stream of work theoretically and empirically. Theoretically, Lemke adds by re-defining states and state death (chapters 1 and 2) and deriving testable implications from state making theories (chapter 3). The conventional definition of states is governing institutions supervising populated territories. To operationalize this, the standard data collection by the Correlates of War (COW) project established two criteria for states: population above 500,000 people (until 1920) and diplomatic recognition (meaning United Nations/League of Nations membership after 1920, and recognition by France and Britain before then). The resulting list of 244 states has become the industry standard and has been widely used. It has also been criticized for inconsistencies and for using the political decision of diplomatic recognition as a coding criterion (also see Fazal 2007; Bridget Coggins, Power Politics and State Formation in the Twentieth Century, 2014).

Lemke’s state definition differs from the conventional COW definition by removing the two criteria: there is no minimum population size and no need for diplomatic recognition by other states. Instead, Lemke argues that “wherever a set of institutions enjoys exclusive territorial control, that is a state” (p. 33). Lemke notes that institutions encompass four aspects: “territorial political entities that govern people, tax them, have coercive and military means to provide internal order and external defense, and make decisions about who benefits from the resources the state has at its disposal” (p. 4).

Lemke’s conceptual innovation is that “states” come in two types: “sovereign states” (those with diplomatic recognition, listed in COW) and “territorial contenders” (state-like entities lacking diplomatic recognition, missing from COW). Territorial contenders (TCs) can be “parts of a sovereign state’s territory that are beyond the control of that state’s institutions” (p. 32). TCs may be found in, for example, “rebel-held territory, government no-go zones, states within a state” (p. 38). Lemke’s re-definition of “states” has two primary benefits: it acknowledges state-like entities that have (as an aggregate group) often been a blind spot in quantitative analyses, and it allows analyzing how state-like entities differ from sovereign states.

In a second move, Lemke expands the definition of state death by noting that territorial contenders can also die, and that territorial contenders can cause a sovereign state to die (see expanded typology in Figure 2.1). This differs from the conventional definition of state death by Fazal (and other scholars) which emphasizes the loss of legal status of a sovereign state and take over by another sovereign state. Due to expanding the scope of states (adding territorial contenders) and types of death (adding internal collapse), the number of state-deaths increases. The conventional death count (by Fazal) is 66 cases of state death in 207 sovereign states between 1816 and 2000. Lemke, in contrast, counts 164 state deaths in 62 sovereign states (or a total of 249 “states” when we account for 187 territorial contenders in these 62 sovereign states).

Empirically, the book contributes a dataset on territorial contenders and state death in 62 sovereign states’ territories in the developing world from 1810 (or the year of sovereign state independence) to 2010. The new dataset provides information on TC birth and death (how TCs began and ceased to control territory); their power (size of their territory, population, military forces; whether they received foreign military support or domestic illicit income); regime type (democracy/mixed/autocracy); and whether/how many sovereign states gave diplomatic recognition. The book’s appendix lists 62 sovereign states and their TCs with their life spans. A great addition here is the online code sheet for each of the 187 TCs, which provides sources and rationales for each entry (totaling 1000 pages of text)—making it easier to understand, check, and change the coding if so desired. This is a model for data transparency and replicability. The book transparently discusses coding decisions, challenges of measurement, difficulties in inference, and attributing causality in observational data.

The book tests whether existing explanations for state death also hold for TCs (chapters 3–6) and largely confirms previous insights. Chapter 3 shows that ongoing conflict is associated with lower risk and defeat with higher risk of death, while victory in war does not shape death risks. Lemke sees the victory null finding as puzzling (pp. 114–117). However, this could be explained by state-making literature (Luis Schenoni, Bringing War Back In: Victory, Defeat, and the State in Nineteenth-Century Latin America, 2024, p. 104), which expects victory to consolidate (not change) trends in a “preparation for war” phase, which may already be picked up by the ongoing conflict variable. The null finding on victory is therefore consistent with accounts of state making.

Chapter 4 tests the norm against violent conquest after World War II. While previous work supports this, Lemke (focusing on the developing world) only finds support in the case of territorial contenders but not sovereign states. This is an interesting issue for future research. Consistent with previous studies, chapter 5 shows that buffer states lead dangerous lives and chapter 6 shows that state strength is not related to state death. Finally, direct comparison shows that territorial contenders are weaker and more likely to die than sovereign states.

One intended contribution of the book is to generate a new list of “states” that can inform future analyses (also see Bennett, Robert and Joseph Zitomersky, “The Delimitation of International Diplomatic Systems: 1816–1970,” in Joseph Zitomersky, ed., On Making Use of History, 1982; Fazal 2004, 2007; Griffiths, Ryan and Charles Butcher, “Introducing version 2 of the International System(s) Dataset,” International Interactions, 46(2), 2020; Paul Diehl et al., “Conceptualizing and Operationalizing the Interstate System,” International Studies Review, 2025). Yet, while the concept of a state as “territorial contender” is new, its manifestations overlap largely with familiar nonstate actors. The territorial contender data cover 84% of rebel groups, 94% of de facto states, and 50% of self-determination/secession movements in commonly used datasets, as such actors are TCs if they control territory (pp. 43, 48). It is not clear what percent of TC observations are “new” actors which are not included in such datasets.

Analytically, perhaps the trickiest aspects of the book are 1) distinguishing the state from rulers/parties/administration—what Max Weber called the “community” (Gemeinschaft) that holds the claim to the monopoly of violence—and 2) distinguishing state death from regime change. On the first, Lemke notes that the state as a governing institution controlling territory is broader than rulers or parties (p. 6). But exactly how much institutionalization and how many governing functions this encapsulates seems to change a bit throughout the book. For example, Lemke also notes that “governing institution” can mean “rebel group” (p. 57) or criminal organization (if they control some territory, p. 36). And that exiling rulers/authorities is state death (p. 7). If so, how does institution differ from a set of individuals? Could it just be a small leadership group controlling some territory, without the institutional infrastructure? And what is the minimum duration that a group must hold territory in order to be a TC?

On the issue of state death versus regime change, Lemke acknowledges that they are sometimes difficult to distinguish (pp. 65–70): not all regime changes are state death; the fine line is territorial control and governing institutions being reorganized rather than replaced. This can be somewhat subjective. For example, when Syria’s Assad regime lost territorial control to opposition forces in late 2024, was that state death or just regime change? Did the main four institutions get replaced, reformed, or reorganized? More generally, how many institutions must change for a state to be dead? This merits more elaboration.

The book inspires future research both for data and analysis. An obvious next step is expanding the TC data beyond 62 sovereign states’ territories. This would also help with inference, as the relative rarity of state death and the rather small sample of 62 sovereign states is sometimes insufficient for models to converge and yield estimates for quantities of interest (p. 228). Future work may also help identify more TCs in the already catalogued 62 states. One concern is that we are more likely to know about TCs when there was a civil war or other conflict. Once conflicts fall below a certain severity threshold, they are less likely to be noted in the media and conflict datasets (von Borzyskowski, Inken and Michael Wahman, “Systematic measurement error in election violence data: Causes and consequences,” British Journal of Political Science, 51(1), 2021).

The book also generates interesting questions for analysis. On the basic difference between TCs and sovereign states, is it the recognition itself that drives state survival or is it that the factors leading to recognition also contribute to state survival (or both)? Also, (how much) do results differ when the same models are run on other non-state actor datasets like rebels (David Cunningham et al., “Non-state actors in civil wars,” Conflict Management and Peace Science, 30(5), 2014), secessionist movements (Cunningham, Kathleen, Inside the Politics of Self-Determination, 2014) and de facto states (Florea, Adrian, “De facto states: Survival and disappearance (1945–2011),” International Studies Quarterly, 61(2), 2017)? Furthermore, what makes some TCs win or turn into sovereign states, while most do not? Is it the usual (power/alliances) or are there other reasons? And would this be different from asking why certain rebel groups win civil wars?

Ultimately, this thought provoking book invites us to see more similarities between domestic and international affairs—a path that other studies are likely to help tread. Lemke’s thinking about “the state” also pushes political science closer to historical sociology, economic history, and the humanities. The book should thus attract great interest from those working at the intersection of IR and domestic affairs in various disciplines, those interested in conflict studies (civil wars, secession movements, regime change, revolutions, coups), and data scientists in these fields.