How leader's type shapes the effect of nuclear latency on dispute involvement

Abstract

What is the effect of nuclear latency on states’ conflict behaviors? Defined as the ability to produce fissile material without the actual possession of nuclear weapons, nuclear latency is much more common in the international system than nuclear weapons are. Yet present studies might have underestimated its impact on conflict. We argue that, given its ambiguous nature as a dual-use technology, nuclear latency has the potential to significantly alter the effect of leader type on conflict. In particular, while conventional wisdom about hawks and doves suggests that the former should be more prone to conflict than the latter, nuclear latency blurs the distinction between leader's types. Bridging together studies of nuclear latency and leader type, we study how nuclear latency shapes the impact of leader type on the propensity of initiating conflict or being targeted. We find that in the absence of latency, hawkish leaders are more likely to initiate disputes than dovish leaders. However, latency muddies the waters. Thus, leader characteristics does not affect dispute initiation for latent states.

Keywords

Conflict domestic politics leaders nuclear latency

The study of nuclear latency is perhaps one of the most interesting developments in the field of nuclear weapons. Defined as states’ ability to produce fissionable material that can be used in the production of both nuclear energy and nuclear weapons, latency is an inherently ambiguous technology. Achieving latent nuclear status matters, in theory, because possessing the requisite capabilities that define latency allows states to reap some of the benefits of nuclear weapons without fully proliferating. Latency is also more widespread than nuclear weapons acquisition. Between 1939 and 2013, 31 states have been latent at some point, while only 13 states have ever deployed nuclear weapons.¹ Recent research has examined latency's impact on multiple processes, including negotiation and broader cooperation, but with an emphasis on international conflict.²

While presenting an interesting and multifaceted overview of the impact of latency in international politics, the literature on nuclear latency operates primarily from the perspective of latency as opportunity. In the case of conflict initiation, for example, being almost nuclear provides an opportunity to use one's latent nuclear capability to alter the status quo. For deterrence, possessing fissionable material provides the opportunity to follow through on the threat at the heart of virtual deterrence, namely, the threat to proliferate. Studies adopting this opportunity-based approach have produced only mixed results about why and how latency matters for conflict. One reason for those mixed results might be that latency is an ambiguous status in terms of its relationship to conflict. Latency is consistent with both the intent to develop civilian nuclear programs and the pursuit of nuclear weapons. Thus, the nature of opportunity here is perhaps less clear than currently assumed. For some, possessing latent capabilities may indeed present an opportunity to alter their international relations in line with the new-found capabilities. For others, however, latency may mean something far more benign.

An important implication of the ambiguous nature of latency is that opportunity is important, but perhaps an incomplete explanation for latency and conflict behavior. We argue in this paper that we can gain additional theoretical traction on the issue of latency and international conflict by theorizing about both opportunity and willingness. In particular, we focus on leaders’ hawkish or dovish predispositions as a predictor of states’ willingness vis-à-vis their latency-based opportunities. We do so for two reasons. First, studies find that leaders are important predictors of the fate of states’ nuclear plans. Nuclear plans are often shrouded in secrecy, which means only a small number of individuals are aware of them. In this context, studies find, leaders have historically played a very big role (Fuhrmann and Horowitz, 2015; Whitlark, 2017).³ Second, studying how leaders’ dovish and hawkish proclivities interact with latency can uniquely enrich the discussions in this special issue on leaders’ impact in international politics. While the study of leaders’ impact on the world stage has mostly focused on conflict (Carter, 2024a), we know less about how leaders’ preferences interact with nuclear latency in this regard. To boot, most investigations have looked at how institutions constrain or enhance leaders’ preferences (see Carter, 2024b; Chiozza and Khalifa, 2024; Greene and Licht, 2024). Here, we look at how technology and the inherent uncertainty linked to it might instead shape the impact of leaders’ preferences.

We argue that leader's type and latency interact to determine conflict initiation and targeting. In particular, we expect latency to shape the effect of leader type on these phenomena. Thus we allow for the possibility that latency means different things as preferences for using those means vary. We provide quantitative tests of our hypotheses using existing data on a country's latency status and leader-year preferences from 1945 to 2004. We find that in the absence of latency, hawkish leaders are more likely to initiate disputes than dovish leaders. However, latency muddies the waters. Thus, leader characteristics do not affect dispute initiation for latent states.

Latency and conflict

Scholars have recently started to explore the role of latency on conflict. Fuhrmann and Tkach (2015) argue that latency could either contribute to deterrence or make conflict against the state with latent capabilities more likely. These claims produce two hypotheses, one that claims that latent states are less likely to be targeted (deterrence) and one that claims latent states are more likely to be targeted by opposing states than are non-latent states. The empirical tests support the deterrence hypothesis. Building on this, Mehta and Whitlark (2017) expand the possible set of behaviors that can be linked with nuclear latency. Using a different research design, Mehta and Whitlark find no evidence that latency provides a deterrent effect. Using violent disputes as their dependent variable, the authors find no significant relationship between latency and being the target of a militarized dispute. Two other results bear mentioning. First, Mehta and Whitlark find that latent states are not only failing to achieve a successful deterrent, they are also more likely to have sanctions placed on them by the United States. This is further evidence that latency brings few positive outcomes for states. Second, the authors also show that latent nuclear states are emboldened by their status, as latency is positively associated with militarized interstate dispute (MID) initiation.⁴

Studies of the effect of latency on conflict have their roots in the broader literature on the effect of nuclear weapons on conflict. As the Cold War deepened and the number of nuclear-armed states expanded beyond the United States and Soviet Union, conflict scholars began the daunting process of theorizing in the shadow of the nuclear revolution. Not surprisingly, much research in this era asked questions about how the emergence of a weapon with the destructive capacity of the atomic, and later the hydrogen, bomb might alter the conflict dynamics between states. Over time work in this area focused on three prominent theoretical narratives that link the possession of nuclear weapons to the use of military force. Early theoretical studies of the nuclear revolution emphasized the deterrent role that nuclear weapons play, and this camp, known as nuclear optimists, argued that nuclear weapons suppress the likelihood of conflict owing to the threat of nuclear retaliation. A second, related, perspective—the stability–instability paradox—suggests both a deterrent and an emboldenment effect (Jervis et al., 1989; Kapur, 2005). According to this perspective, nuclear parity in a dyad may produce a deterrent effect at the level of nuclear war, while encouraging lower-level conventional conflict precisely because escalation to nuclear war was stifled by a fear of nuclear retaliation. Finally, a third position, that held by so-called nuclear pessimists, embraces the conflict element of the paradox. Nuclear pessimists fear that the acquisition of nuclear weapons by a state may increase the chances for military conflict either because the nuclear-armed state is emboldened by its new capabilities and becomes militarily aggressive as a result, or because a state believed to be going nuclear may be the target of a preventive attack before proliferation can occur.⁵

Many quantitative studies investigate and draw conclusions about all three of these arguments, and the results for all three theoretical positions are decidedly mixed. Some find support for the nuclear pessimism,⁶ others for nuclear optimism.⁷

Latency as opportunity

Can latency have a similar effect on conflict to that of nuclear weapons? We expect this not to be the case. Latency itself is a much more ambiguous status than the possession of nuclear weapons is. States that possess ENR facilities can build nuclear weapons more easily than those states that do not possess those facilities. But those facilities do not offer access to the option of using nuclear weapons in conflict as readily as the actual possession of a nuclear arsenal, should states even wish to do. For this reason, compared with the actual possession of nuclear weapons, latency represents a much noisier signal of whether states will use nuclear weapons in conflict in the future.

Indeed, studies find that nuclear latency has a complex effect on a state's nuclear agenda. One the one hand, it incentivizes states to establish a nuclear weapons program and even to acquire nuclear weapons. On the other hand, it increases the probability that states will abandon an existing program. Nuclear latency's complex effect is produced by the fact that dual-use technology reduces the costs of both starting a program and restarting it once it has been paused (Mattiacci and Jones, 2016). We argue that latency does not affect states’ propensity for conflict the same way. Instead, its impact depends on leader preferences.

Latency and emboldenment

Nuclear latency is an important scientific, and potentially military, achievement that enhances the capacity of a state. Given this, it is natural to follow Bell's lead and conclude that states use their advanced capability to pursue a more aggressive approach to international relations precisely because they expect to be more successful in revising the status quo as the result of the benefits that latency and the threat of proliferation afford. Bell addresses this explicitly in his typology on nuclear behaviors and this claim is consistent with with his third assumption about states using their power to address their interests. He argues that possessing nuclear weapons may make using force more attractive, in part, because the nuclear option reduces the costs of coercion by providing “a shield behind which aggression can be taken.”

Latency and virtual deterrence

We can also think about these issues within the virtual deterrence context. Given the prominence of the nuclear revolution literature and its emphasis on deterrence, it is not surprising to see research hypothesizing that latency might be linked with deterrence benefits. While nuclear deterrence rests on the threat of a credible use of nuclear weapons by the target of an attack, virtual deterrence rests on the assumption that the potential attacker fears proliferation and thus acts with caution. Virtual deterrence thus rests on the credibility of the latent state's threat to proliferate. If a latent state is attacked, it could decide that the best way to ensure future deterrence is to move from latency to the full acquisition and deployment of nuclear weapons. Given that proliferation and the resulting change in the bargaining situation (Sobek et al., 2012) between the two states are not in the long-term interests of the potential attacking state, caution is induced and deterrence achieved. Latency again creates an opportunity for states to threaten proliferation in a meaningful way. Constructing and deploying even a small nuclear arsenal is much more likely for a state that has already achieved latency than for a state that lacks the ENR facilities.

Together, these opportunity arguments suggest two baseline expectations about latency and conflict that mirror those expectations offered in the extant literature. First, nuclear latent states should be more likely to initiate conflicts than non-latent states. Second, latent states should not be targeted as often as states that lack latent status. In both cases the expected effect is produced by the opportunity afforded by being almost nuclear. Below, we move beyond opportunity to consider how a leader's relative hawkishness or dovishness works in the context of nuclear latency.

H1: Nuclear latent states are more likely to initiate disputes than non-nuclear latent states, regardless of leader type.

H2: Nuclear latent states are less likely to be targeted in disputes than non-nuclear latent states, regardless of leader type.

Leader preferences, nuclear latency, and conflict behavior

Whether thinking about nuclear latency and emboldenment or latent capabilities and virtual deterrence, the literature to date has assumed that latency is fundamentally a technological condition that has a uniform effect on conflict. Once the technological hurdles involved in developing ENR facilities have been overcome, the benefits of latency follow. Mehta and Whitlark (2017) suggest that gaining ENR capacity produces a credible threat to proliferate. Volpe (2017) likewise argues that credibility is linked to technological capacity. He argues that states seeking to use their latent status to compel a rival face a credibility dilemma that is overcome when the state finds a technological “Goldilocks Zone”, defined as going far enough towards nuclear acquisition to credibly proliferate, but not so far as to prompt strong counterproliferation actions by others who fear proliferation. This focus on the opportunity to go nuclear by virtue of having the capacity is, of course, a natural place to begin theorizing about the effects of latency on conflict, as possessing the technological ability to go nuclear is a necessary condition for any latency effect. However, existing studies on latency and conflict limit their theoretical expectations by assuming that all states with ENR facilities are equally likely to go further with their nuclear weapons programs.

A next step, then, is to think about the possibility that latency means different things in different political contexts. For some states achieving nuclear latency may be seen by their rivals as menacing. For other states, though, latency may not move the needle on the probability of conflict. As Levite (2003) argues, latency is often a byproduct of processes that, while consistent with pursuit of nuclear weapons, have more to do with non-military outcomes. As captured by the grand bargain in the Treaty on the Nonproliferation of Nuclear Weapons, for example, the international community has long advocated that states develop peaceful uses of nuclear power. Levite's point here meshes nicely with the idea that latency is best understood if we consider more than just the technological aspect of the issue. To her, the missing element is what she deems “nuclear ambivalence.” Consider two states that possess nuclear latency, but where leaders with different levels of nuclear ambivalence are in positions to make key decisions about nuclear weapons. In the case where nuclear ambivalence is high, latency is likely to be less threatening than in a state where the leader has low ambivalence towards a weapons program.

Nuclear ambivalence, however, is hard to capture systematically. We argue that a good way to get at this concept is through the related notion of leaders as hawks and doves. That hawks and doves approach international relations differently is well established in the literature, and even a brief overview of how hawks and doves are characterized will demonstrate why preferences matter in the context of nuclear latency. First, and most generally, hawks are often portrayed as realists while doves are identified as idealists (Russett, 1990). Not surprisingly, anchoring leader preferences in these broad theoretical traditions produces many attitudinal and behavioral expectations. The most prominent theme in the international relations literature is the research agenda that links hawks and doves with different propensities for engaging in military conflict. In their recent work, for example, Mattes and Weeks (2019: 58) explicitly identify hawks and doves by their relative willingness to use military force. They describe a hawk as a leader who has “a reputation for favoring military solutions over diplomatic ones,” while doves are the opposite in terms of their preferred policy options . In another recent piece that utilizes a hawk/dove approach, Heffington (2018) offers a quantitative approach to identifying hawks and doves. Specifically, he builds his militarization index by identifying content from a leader's party manifesto. According to Heffington, hawks are those leaders whose party platforms contain a large number of pro-military statements while also presenting few statements emphasizing peace or otherwise suggesting a de-emphasis on the military.

A second, related factor attributed to hawks is that they also subscribe strongly to the importance of deterrence. Russett (1990) raised this issue, noting that hawks are unlikely to engage in significant levels of disarmament given that they view the world through a realist lens. As a result, hawks rely on deterrence to maintain the status quo. This idea is also included in Weeks and Mattes’ definition of a hawkish leader that was discussed immediately above. The fictional American president used in their experiment is not only described as preferring the military option, but their summary of his positions also details that he believes that security is achieved through deterrence. Kesgin (2020) attempts to classify hawks and doves, and does so with Russett's initial point on deterrence in mind. Because hawks are highly distrustful of others, Kesgin posits, they are wary of their adversaries and thus more likely to accept deterrence as a key policy position. That hawks lack trust while doves do not is also relevant for the final way that hawks and doves are typically identified, and in this case the issue tied how leaders approach international cooperation. Rathbun (2011) extends the logic of leader types to preferences for how states engage with other states. He suggests that doves are more likely to seek out opportunities for multilateral cooperation than are hawks, who place greater value on unilateral actions. Consistent with this, Rathbun also argues that hawks are predisposed to defect in the face of multilateral cooperation, while dovish leaders seek to maximize their states’ compliance with the rules that govern cooperation.

Linking leader preferences and latency to conflict

While reaching different answers as to what the ultimate effect of latency is, existing studies argue that latency affects conflict systematically and unconditionally. We now turn to thinking about how leader preferences might interact with latency to produce different expectations. We apply our argument to both a state's decision to initiate a MID (emboldenment) and also the question of whether latent states will be the targets of military aggression (virtual deterrence).

The conditional effect of leader type on MID initiation

The literature on leader type and conflict has found a clear and consistent pattern of behavior: hawkish leaders are more likely to initiate conflicts than are dovish leaders. However, the acquisition of nuclear latency may scramble this conventional wisdom by altering the international environment that both hawks and doves confront. As a result, doves may face incentives to behave in a more aggressive manner than they otherwise might, whereas hawks will face new incentives for restraint. To understand this shift in incentives, it is important to consider the consequences of the acquisition of nuclear latency. In particular, there is the notion that nuclear latency may embolden states. Nuclear latency enables it to more credibly threaten to acquire nuclear weapons if threatened. As a result, latency might incentivize states to pursue a more aggressive posture, precisely because they may expect to be more successful in disputes as a result of the benefits that latency, and the implicit threat of nuclear acquisition, afford.⁸ Especially relevant here, Mehta and Whitlark (2017) adopt this framework in the context of nuclear latency. Invoking the logic of the stability–instability paradox, they suggest that a latent nuclear deterrent may encourage states to initiate conflicts they would not have started absent their latent capability.

While nuclear latency is posited to increase incentives for aggression, hawks and doves may respond somewhat differently to these capabilities. Hawks, we argue, are more likely to be low on nuclear ambivalence and thus more willing to use a latency advantage to press for gains on their interests abroad. Given their predisposition towards aggressive foreign policy, hawks are likely to see latency as an enhancement of their preference for coercion and using force. Doves, on the other hand, are less likely to see latency as an opportunity. Their fundamental preference for cooperation over conflict and the lack of motivation to follow through on a threat to proliferate suggests that latency will have no discernible effect on the international behavior of states with dovish leaders.

H3: Among nuclear latent states, hawks are more likely to initiate disputes than doves.

The conditional effect of leader type on MID targeting

Existing studies of virtual deterrence assume that latency works as a deterrent force for all types of leaders. Given that virtual deterrence is driven in large part by the potential attacker's fear of nuclear proliferation by the target state, however, it is possible that hawks and doves may derive different benefits from virtual deterrence. While possessing the technical capacity is a requirement for proliferation, having the willingness to proliferate also matters for credibility. In other words, credibility varies across deterrence contexts (Zagare et al., 2000), where context is determined by leader preferences. A potential attacker is more likely to be deterred by a latent hawk than a latent dove.

Why might hawkish leaders be seen as more credible in the context of virtual deterrence? Hawks are more distrustful of others than are dovish leaders. As such, hawks emphasize security, deterrence, and military solutions to international problems. All of this points toward an increased willingness to proliferate if attacked by a rival because of the perceived deterrent value of deployed nuclear weapons in the future. Hawks playing to type will be determined to ward off additional future attacks by following through on their threat to go nuclear. Forward-thinking hawks may be especially inclined to go nuclear because the decision to do so, once made, “locks in” nuclear status and takes the choice away from future dovish leaders.⁹

Moreover, hawks may have greater credibility with respect to virtual deterrence because they may be more willing to violate international norms against proliferation than doves. Nuclear weapons are regulated by an extensive set of international institutions whose purposes include monitoring and verification of the regime's rules. The linchpin of the regime is the Treaty on the Nonproliferation of Nuclear Weapons (NPT), but hawks and doves may perceive the international constraints and the threat of punishment stemming from this regime differently. As noted by Fuhrmann and Lupu (2016), NPT pessimists see the treaty as a relatively weak institution given the inconsistency in efforts to enforce it, while others (NPT optimists) see the treaty as a significant barrier for proliferators to overcome. Overall, there is some evidence that states are willing to enforce the nuclear regime. Coe and Vaynman (2015), for example, find evidence the United States and Soviet Union have worked together to prevent proliferation. Others have shown that military strikes (Fuhrmann and Kreps, 2010; Kreps and Fuhrmann, 2011) and sanctions (Miller, 2014) have worked at various points during the post-1945 period. Nevertheless, hawks may be more likely than doves to interpret the inconsistency in enforcement as an opening to acquire nuclear weapons.

H4: Among nuclear latent states, hawkish leaders are less likely to be targeted than dovish leaders

Research design

To test our theory, we engage in a comprehensive study of all states in the international system from 1939 to 2004. Countries enter the dataset in 1939, or in the first year they gain independence thereafter. All countries remain in the dataset until 2004 or until they stop existing, if that happens earlier. The unit of analysis is the country-year.

Dependent variables

Using the Militarized Interstate Disputes (MID) v.5.0 dataset (Palmer et al., 2022), we code two dependent variables using the approach of Prins (2001), among others. First, we code one variable as 1 if a leader is the initiator of a MID onset in a given year, and zero otherwise. Second, to capture whether a leader is targeted, we code a variable 1 if a leader is targeted in a MID onset in a given year, and zero otherwise.¹⁰ For both of these variables, our focus is on the states involved at the time the dispute is initiated, so joiners are coded as 0. Similarly, our focus is on dispute onset, so all ongoing dispute years are coded as missing.

Independent variables

To measure a leader's preferences with respect to the use of force, we employ a latent measure of leader hawkishness developed by Carter and Smith Jr (2020). Lower values of this continuous variable correspond to more dovish leader preferences, and higher values correspond to more hawkish leader preferences. For country-years with more than one leader, we average each of the leaders’ hawkishness values.

We use the measure developed by Carter and Smith Jr (2020) because it is the most comprehensive measure available. As Carter and Smith Jr (2020) note, there are several different measures of leaders’ hawkishness. Most of them focus, somewhat arbitrarily, on a single specific leader characteristic as a measure of their hawkish preferences. They rely, for example on the leader's psychological features (e.g. Saunders, 2011), their background experiences (e.g. Horowitz et al., 2015), or their political ideology (e.g. Bertoli et al., 2019). By exclusively focusing on one characteristic, Carter and Smith Jr (2020) demonstrate, researchers risk overlooking important determinants of a leader's preference for the use of force. By employing a Bayesian latent variable approach, Carter and Smith Jr (2020) are able to leverage information from each of these different traditions to generate a more holistic and accurate measure of leader hawkishness, reflecting the fact that this underlying concept is multidimensional.

Using the Nuclear Latency Dataset (Fuhrmann and Tkach, 2015), we measure nuclear latency with a dichotomous indicator that acquires the value 1 if in a particular leader-year the leader's country operates one or more ENR facilities. While these facilities might vary in size or purpose, they all reflect the countries’ intention and capacity to produce fissile material, which is the concept we want to measure. Because the concept we are interested in is a state's latent nuclear capacity, once a state acquires nuclear weapons, they are no longer coded as possessing nuclear latency, even though they may continue to operate ENR facilities in their territory.¹¹

Control variables

We include a number of control variables to capture processes that might confound the relationship between each leader's political preferences, their state's propensity to become involved in a militarized dispute, and nuclear latency.

To this end, we control for three categories of factors that might have this effect. First, we control for factors relating to a state's nuclear technology. Thus, we include a dichotomous measure indicating whether one or more of a state's ENR facilities is subject to International Atomic Energy Agency (IAEA) controls in a given year. Then, we include a dichotomous measure indicating whether a state operates one or more ENR facilities while engaging in an attempt to keep these ENR facilities covert a given year. Each of these measures might potentially impact the relationship between nuclear latency and disputes by making both less likely. To operationalize each of these variables, we rely on Fuhrmann and Tkach (2015). Moreover, we include a dichotomous measure indicating whether a state has acquired nuclear weapons, as such states are no longer latent but may nevertheless exhibit different conflict behavior from states that are not latent and do not possess nuclear weapons.¹²

Second, we control for international factors. Specifically, we control for whether a state has received a nuclear deployment from another nuclear power using data from Fuhrmann and Sechser (2014). Nuclear deployments, Fuhrmann and Sechser (2014) find, correlate with the decision to protect those states where the nuclear deployment is located. This decision to protect allies, in turn, might confound the relation between dispute initiation and latency by making all three phenomena more likely. Moreover, we control for the prevalence of the nuclear non-proliferation norm in the international system, as the growing strength of this norm may alter leaders’ perceptions about the utility of leveraging nuclear latency in periods of conflict. To operationalize this norm, we generate a measure of the proportion of states in the international system that have ratified the NPT in each year. Finally, we control for whether a state has at least one rival with either latency or nuclear weapons, using rivalry data from Thompson and Dreyer (2010). Rivals with some measure of nuclear capacity may increase both the likelihood of a state pursuing latency and the likelihood of being involved in a dispute.

Third, we control for domestic factors. Thus, we control for a country's regime type using the combined polity score measure from the Polity 4 dataset (Marshall and Jaggers, 2002). Regime type might confound the relationship between leader preferences and dispute behavior as it is possible that certain regime types might be more likely to select more hawkish/dovish leaders and also have systematically different dispute behaviors than others. Second, we add a variable that captures a country's gross domestic product (GDP) per capita (Penn World Tables, Feenstra et al., 2015), as a country's wealth might impact the type of leader it is likely to select as well as its propensity to initiate and be targeted in disputes.

Results

Table 1 presents results from two models where an interaction between nuclear latency and leader type predicts whether a country in a given year is more likely to initiate a dispute (Model 1) or be targeted in a dispute (Model 2). Given the presence of interactions to test our main variables, however, it is best to interpret the results by estimating predicted probabilities.¹³

Table 1.

The effect of latency and leader preference on conflict.

	Model 1	Model 2
	MID initiation	MID targeting
Leader type	0.150*	0.134
	(−1.86, 0.303) (0.017, 0.280)	(−0.044, 0.322) (−0.013, 0.296)
Latency	0.847**	0.676
	(0.045, 1.65) (0.173, 1.510)	(−0.193, 1.55) (−0.06, 1.43)
Leader × Latency	−0.1	0.0974
	(−0.518, 0.352) (−0.452, 0.294)	(−0.514, 0.763) (−0.397, 0.618)
IAEA controls	−0.747	−1.13**
	(−1.76, 0.352) (−1.64, 0.138)	(−2.2, −0.063) (−1.97, −0.296)
Rival	0.59**	0.621**
	(0.298, 0.879) (0.342, 0.85)	(0.316, 0.951) (0.366, 0.895)
Covert facilities	0.0845	0.208
	(−0.554, 0.748) (−0.453, 0.639)	(−0.456, 0.875) (−0.36, 0.781)
Regime type	−0.000	0.002
	(−0.019, 0.019) (−0.017, 0.017)	(−0.020, 0.024) (−0.018, 0.0215)
Percentage of states in NPT	0.38**	−0.0789
	(0.011) (0.076, 0.686)	(−0.433, 0.278) (−0.383, 0.231)
GDP pc	0.000	−0.000
	(−0.000, 0.000) (−0.000, 0.000)	(−0.000, 0.000) (−0.000, 0.000)
Nuclear deployments	0.0987	0.136
	(−0.444, 0.646) (−0.336, 0.563)	(−0.316, 0.596) (−0.241, 0.53)
Acquisition	1.12**	0.797**
	(0.611) (0.703, 1.56)	(0.417, 1.19) (0.4771, 1.1)
$t$	−0.256**	−0.28**
	(−0.316, −0.197) (−0.305, −0.206)	(−0.356, −0.201) (−0.342, −0.214)
$t^{2}$	0.009**	0.012**
	(0.005, 0.013) (0.005, 0.012)	(0.007, 0.0170) (0.008, 0.016)
$t^{3}$	−0.000**	−0.000**
	(−0.000, −0.000) (−0.000, −0.000))	(−0.000, −0.000) (−0.000, −0.000)
Constant	−1.51	−1.79
	(−1.86, −1.18) (−1.8, −1.22)	(−2.15, −1.4) (−2.1, −1.48)
N	4746	4746
AIC	3134.439	1943.721

*p ≤ 0.10; **p ≤ 0.05.

Standard errors, clustered by country. 95% confidence intervals, followed by 90% confidence intervals, reported in parentheses beneath mean coefficient estimates.

AIC, Akaike information criteria; GDP pc, Gross domestic product per capita; IAEA, International Atomic Energy Agency; NPT, Treaty on the Nonproliferation of Nuclear Weapons.

Table 2 presents substantive estimates from the results of Model 1. The left side of the table estimates the change in the probability of a state initiating a MID associated with going from a dovish leader to a hawkish leader, varying whether a state possesses latency or not.¹⁴ For states without nuclear latency, we find, consistent with previous studies of leader preference and conflict, that hawkish leaders are more likely to initiate a dispute than dovish leaders. The change in the probability of initiating a dispute is positive and statistically significant at the 90% level. Interestingly, and counter to H3, we find that leader type no longer affects the probability of conflict initiation once a state possesses nuclear latency, as the change in the probability of MID initiation from doves to hawks is not statistically significant. The results suggest that the uncertainty surrounding the effect of leaders on the risk of MID initiation in the context of latent states is quite substantial.¹⁵

Table 2.

Substantive effects, militarized interstate dispute (MID) initiation.

	Effect of leader type			Effect of latency
	No latency	Latency		Dove	Hawk
Dove $\to$ Hawk	0.0173*	0.008	No latency $\to$ Latency	0.115**	0.104*
	(−0.002, 0.037)	(−0.073, 0.100)		(0.002, 0.295)	(−0.006, 0.283)
	(0.002, 0.034)	(−0.063, 0.083)		(0.012, 0.256)	(0.010, 0.235)

*p ≤ 0.10; **p ≤ 0.05.

95% confidence intervals presented beneath mean estimates, followed by 90% confidence intervals.

The right side of Table 2 estimates the change in the probability of a state initiating a MID associated with a state going from not having latency to having latency, varying whether the leader of the state is a dove or a hawk. These results suggest that a country with nuclear latency is more likely to initiate a MID than a country with no nuclear latency. This is the case whether the leader is dovish or hawkish. In this sense, latency blurs the distinction between leader's type when it comes to conflict preference: latency has a similar, positive and statistically significant effect on conflict initiation, regardless of whether hawks or doves are in power. This finding is consistent with H1.

Our results for MID targeting (Model 2) differ from our results with respect to MID initiation. Broadly speaking, we find little evidence of an interactive relationship between leader type and latency, on the one hand, and a country's risk of being targeted in a dispute on the other. Re-estimating the substantive effects presented in Table 2 with respect to the probability of being targeted in a dispute reveals that none of the effects achieve statistical significance. A country moving from a dovish to a hawkish leader, all else equal, has no significant effect on its probability of being targeted, regardless of whether it possesses latency or not. Similarly, a state acquiring latency has no significant effect on the probability of it being targeted, regardless of whether it possesses latency or not.¹⁶

Discussion

We find support for H1: nuclear latent states are more likely to initiate disputes than non-nuclear latent states, regardless of leader type. This result holds across different specifications of the model (see the Online Appendix). We do not find support for H2, as acquiring latency does not affect the likelihood of being targeted. In addition, our results add important nuances to the ways we understand the role of latency and of leader's type on conflict propensity.

Consider the results on conflict initiation. Our conditional expectation was that hawks would be emboldened by latency to a greater degree than doves, even if doves themselves were emboldened to a degree by latency. Our findings indicate that latency blurs the distinction between hawks and doves when it comes to initiating conflict. This result emphasizes the ambiguous state of latency and forces us to rethink the complexities of the impact of ENR facility possession in the international arena.

So, why would latency blur the distinction between hawks and doves? Given the ambiguous nature of the opportunities that latency provides, adversaries may look to the leader's preferences in an effort to estimate what impact latency is likely to have on a state's behavior. Adversaries that see latency as potentially leading to aggression are going to be even more inclined to be threatened by latency when faced with a hawkish leader.¹⁷ Thus, although hawkish leaders may be emboldened by nuclear latency, opportunities to use force may decline as potential targets adjust their policies in light of this particular nuclear context.¹⁸ It is also possible that there is simply a threshold effect for emboldenment and conflict initiation and latency does not affect the level of emboldenment for leaders who are already significantly inclined to initiate conflict. Latency, in other words, increases the probability of hawks initiating, but just does not provide the same magnitude of incentive as it does for less-hawkish leaders.

In contrast to hawks, nuclear latency may greatly increase the incentives for aggression for more dovish leaders, causing them to act in a manner more similar to hawks. While nuclear latency may be perceived to embolden states, other countries may discount this dynamic to some extent when confronting a dovish leader with nuclear latency. As a result, unlike with hawks, other countries may engage in a strategic conflict avoidance to a lesser extent when confronting more dovish leaders. The net result of this dynamic is dovish leaders, emboldened by the acquisition of nuclear latency, confronting an international environment in which other countries will be less inclined to acquiesce to their demands, relative to their response to more hawkish leaders with latency.

Finally, we considered the nature of MIDs as a measure of conflict, as our dependent variable in our main analysis does not distinguish between the different levels of MID severity. It is possible that doves may be emboldened to use force, but only at the lowest levels of severity for conflict initiation. Although not presented above, we ran models in consideration of this possibility. Our findings were similar when utilizing violent MIDs initiation rather than all MIDs initiation, suggesting even in those cases, latency blurs the distinction between hawks and doves, whereas in the absence of latency, hawks are more likely to initiate conflict than doves.

We also expected to find that, given latent status, hawks would be targeted less by adversaries than would doves. Given that virtual deterrence relies on the threat to proliferate, we expected hawks to benefit more from virtual deterrence than doves. In reality, however, doves and hawks are not targeted differently depending on whether they possess nuclear latency or not. This result, again, echoes the fact that latency is an ambiguous signal and that opponents might not know how to interpret it.

Conclusions

Nuclear latency, we have argued, presents states with an opportunity to do things differently in the international arena. Whether states will capitalize on this opportunity or not, in turn, depends on their leaders’ inclinations. Studying the interaction between leader's type and latency and their effect on dispute behavior, we find that latency shapes the impact of leader type. In the absence of latency, hawkish leaders are more likely to initiate disputes than dovish leaders. However, latency muddies the waters. Thus, leader characteristics do not affect dispute initiation for latent states.

Our results have broad implications for both studies of nuclear latency and of leader type, as well as for policymakers. First, these results point to the indirect effect that nuclear latency may have. While scholars so far have tested for a direct effect of latency on conflict, we find that latency shapes incentives via leader type. Latency, therefore, will present different opportunities to different leaders. As a consequence, not all latent states will behave similarly. Given the numerous areas of international politics where the impact of latency has been investigated (Mattiacci et al., 2022), this raises the question: is the impact of leader type on other areas too conditioned by latency? For instance, does latency also shape the impact of leader type on conflict duration and outcome?

Second, our findings speak to the broader debate on leaders’ types. While leaders’ types have been used to predict numerous outcomes in international politics, our results suggest that ambiguous technological advancements such as latency might de facto compress the difference between hawks and doves, making the one group more like the other. Consistent with Carter and Nordstrom (2017), we emphasize that the importance of context becomes central in studying the effect of leader type and conflict. Other studies in this special issues have emphasized how the institutional context can constrain or enhance the impact of leaders’ preferences on a variety of political phenomena (see Carter, 2024b; Chiozza and Khalifa, 2024; Greene and Licht, 2024). Here, we demonstrate instead how technology and the inherent uncertainty often linked to it might make the distinction between those preferences less sharp. Moreover, given the importance of new technology and decision making in the current security context, the importance of the technological context raises the question: what other technological or political factors might have also have this effect?

Finally, these results offer some implications for policymakers engaging in counterproliferation efforts with respect to states such as Iran. Much has been made of the recent presidential transition in Iran, with the hawkish nature of the new president being a frequent topic amongst regional neighbors and those involved in the Joint Comprehensive Plan of Action (JCPOA). The entry of a leader with more hawkish tendencies has possible implications of a new leadership style on negotiations,¹⁹ but our results show that the ambiguity of latency might make the difference between hawkish and dovish leaders less salient in this particular context. If this is the case, then, it is probable that a new leader might not present a radically greater challenge to negotiations, despite any hawkish tendencies.

Supplemental Material

sj-pdf-1-cmp-10.1177_07388942231209952 - Supplemental material for How leader's type shapes the effect of nuclear latency on dispute involvement

Supplemental material, sj-pdf-1-cmp-10.1177_07388942231209952 for How leader's type shapes the effect of nuclear latency on dispute involvement by Benjamin Jones, Eleonora Mattiacci and Timothy Nordstrom in Conflict Management and Peace Science

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Timothy Nordstrom

Supplemental material

Supplemental material for this article is available online.

Notes

References

Bell

(2015) Beyond emboldenment: How acquiring nuclear weapons can change foreign policy. International Security 40(1): 87–119.

Bell

Miller

(2015) Questioning the effect of nuclear weapons on conflict. Journal of Conflict Resolution 59(1): 74–92.

Bennett

Stam

III (2004) The Behavioral Origins of War. Ann Arbor, MI: University of Michigan Press.

Bertoli

Dafoe

Trager

(2019) Is there a war party? Party change, the left–right divide, and international conflict. Journal of Conflict Resolution 63(4): 950–975.

Carter

(2024a) Introduction to special issue: New research on leaders and peace science. Conflict Management and Peace Science 41(2): 99–109.

Carter

(2024b) Political leaders and military spending. Conflict Management and Peace Science 41(2): 132–154.

Carter

Nordstrom

(2017) Term limits, leader preferences, and interstate conflict. International Studies Quarterly 61(3): 721–735.

Carter

Smith

Jr (2020) A framework for measuring Leaders’ willingness to use force. American Political Science Review 114(4): 1352–1358.

Chiozza

Khalifa

(2024) The harsh face of the empire by invitation: Coups in the US world order. Conflict Management and Peace Science 41(2): 110–131.

10.

Coe

Vaynman

(2015) Collusion and the nuclear nonproliferation regime. The Journal of Politics 77(4): 983–997.

11.

Cohen

(2018) Fear and loathing: When nuclear proliferation emboldens. Journal of Global Security Studies 3(1): 56–71.

12.

Early

Asal

(2018) Nuclear weapons, existential threats, and the stability–instability paradox. The Nonproliferation Review 25(3–4): 223–247.

13.

Feenstra

Inklaar

Timmer

(2015) The next generation of the Penn world table. The American Economic Review 105(10): 3150–3182.

14.

Fuhrmann

Horowitz

(2015) When leaders matter: Rebel experience and nuclear proliferation. The Journal of Politics 77(1): 72–87.

15.

Fuhrmann

Kreps

(2010) Targeting nuclear programs in war and peace: A quantitative empirical analysis, 1941–2000. Journal of Conflict Resolution 54(6): 831–859.

16.

Fuhrmann

Lupu

(2016) Do arms control treaties work? Assessing the effectiveness of the nuclear nonproliferation treaty. International Studies Quarterly 60(3): 530–539.

17.

Fuhrmann

Sechser

(2014) Nuclear strategy, nonproliferation, and the causes of foreign nuclear deployments. Journal of Conflict Resolution 58(3): 455–480.

18.

Fuhrmann

Tkach

(2015) Almost nuclear: Introducing the nuclear latency dataset. Conflict Management and Peace Science 32(4): 443–461.

19.

Gartzke

D-J

(2009) Bargaining, nuclear proliferation, and interstate disputes. Journal of Conflict Resolution 53(2): 209–233.

20.

Greene

Licht

(2024) Donor political preferences and the allocation of aid: Patterns in recipient type. Conflict Management and Peace Science 41(2): 155–176.

21.

Hainmueller

Mummolo

(2019) How much should we trust estimates from multiplicative interaction models? Simple tools to improve empirical practice. Political Analysis 27(2): 163–192.

22.

Heffington

(2018) Do hawks and doves deliver? The words and deeds of foreign policy in democracies. Foreign Policy Analysis 14(1): 64–85.

23.

Horowitz

Stam

Ellis

(2015) Why Leaders Fight. New York: Cambridge University Press.

24.

Huth

Bennett

Gelpi

(1992) System uncertainty, risk propensity, and international conflict among the great powers. Journal of Conflict Resolution 36(3): 478–517.

25.

Jervis

(1989) The Meaning of the Nuclear Revolution: Statecraft and the Prospect of Armageddon. Ithaca, NY: Cornell University Press.

26.

Kapur

(2005) India And Pakistan’s unstable peace: Why nuclear South Asia is not like Cold War Europe. International Security 30(2): 127–152.

27.

Kesgin

(2020) Features of foreign policy birds: Israeli prime ministers as hawks and doves. Cooperation and Conflict 55(1): 107–126.

28.

Kreps

Fuhrmann

(2011) Attacking the atom: Does bombing nuclear facilities affect proliferation? Journal of Strategic Studies 34(2): 161–187.

29.

Levite

(2003) Never say never again: Nuclear reversal revisited. International Security 27(3): 59–88.

30.

Liberman

(2001) The rise and fall of the South African bomb. International Security 26(2): 45–86.

31.

Marshall

Jaggers

(2002) Polity IV Project: Political Regim Characteristics and Transitions, 1800–2002. College Park, MD: Integrated Network for Societal Conflict Research, Center for International Development and Conflict Management.

32.

Mattes

Weeks

(2019) Hawks, doves, and peace: An experimental approach. American Journal of Political Science 63(1): 53–66.

33.

Mattiacci

Jones

(2016) (Nuclear) change of plans: What explains nuclear reversals? International Interactions 42(3): 530–558.

34.

Mattiacci

Mehta

Whitlark

(2022) Atomic ambiguity: Event data evidence on nuclear latency and international cooperation. Journal of Conflict Resolution 66(2): 272–296.

35.

Mehta

Whitlark

(2017) The benefits and burdens of nuclear latency. International Studies Quarterly 61(3): 517–528.

36.

Miller

(2014) The secret success of nonproliferation sanctions. International Organization 68(4): 913–944.

37.

Palmer

Mcmanus

Orazio

, et al. (2022) The MID5 Dataset, 2011–2014: Procedures, coding rules, and description. Conflict Management and Peace Science 39(4): 470–482.

38.

Prins

(2001) Domestic politics and interstate disputes: Examining U.S. MID involvement and reciprocation, 1870–1992. International Interactions 26(4): 411–438.

39.

Rathbun

(2011) Trust in International Cooperation: International Security Institutions, Domestic Politics and American Multilateralism, New York, NY: Cambridge University Press.

40.

Rauchhaus

(2009) Evaluating the nuclear peace hypothesis: A quantitative approach. Journal of Conflict Resolution 53(2): 258–277.

41.

Russett

(1990) Doves, hawks, and US public opinion. Political Science Quarterly 105(4): 515–538.

42.

Saunders

(2011) Leaders at War: How Presidents Shape Military Interventions. Ithaca, NY: Cornell University Press.

43.

Saunders

(2019) The domestic politics of nuclear choices—A review essay. International Security 44(2): 146–184.

44.

Smith

Spaniel

(2020) Introducing -Clear: A latent variable approach to measuring nuclear proficiency. Conflict Management and Peace Science 37(2): 232–256.

45.

Sobek

Foster

Robison

(2012) Conventional wisdom? The effect of nuclear proliferation on armed conflict, 1945–2001. International Studies Quarterly 56(1): 149–162.

46.

Souva

Prins

(2006) The liberal peace revisited: The role of democracy, dependence and development in militarized interstate dispute initiation, 1950–1999. International Interactions 32(2): 183–200.

47.

Thompson

Dreyer

(2010) Handbook of International Rivalries, 1494–2010. Washington, DC: CQ Press.

48.

Treier

Jackman

(2008) Democracy as a latent variable. American Journal of Political Science 52(1): 201–217.

49.

Volpe

(2017) Atomic leverage: Compellence with nuclear latency. Security Studies 26(3): 517–544.

50.

Whitlark

(2017) Nuclear beliefs: A leader-focused theory of counter-proliferation. Security Studies 26(4): 545–574.

51.

Zagare

Kilgour

, et al. (2000) Perfect Deterrence. New York, NY: Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.04 MB

0.00 MB