Sage Journals: Discover world-class research

Abstract

The successful closure of arms control negotiations today is conditional on the commitment of many more states than during the Cold War. The question of what determines states’ positions on arms control has therefore become increasingly relevant. Multiple scholars have identified external security threats by other states as the key explanatory factor of opposition to arms control, but empirical evidence hereof is so far limited to a small set of cases. Against this backdrop, this article carries out a global examination of the effect of external threats in the form of interstate disputes and rivalries on state support for arms control. This analysis is facilitated by a novel measure of arms control support that combines United Nations General Assembly voting data with manual coding of 1,178 resolutions. Across a variety of model specifications, the results do not show any significant effect of external threats on support for arms control. This article argues that this means either that the two variables are not related at all, or that two opposing mechanisms cancel out each other: arms control not only entails costs, but also benefits for states that face external threats, as it limits both states involved in a rivalry or dispute. Either way, this study challenges the notion that there is a strictly negative relationship between external threats and arms control support and thus contributes to our understanding of arms control and foreign policy making in general.

Keywords

arms control external threats interstate disputes interstate rivalries United Nations General Assembly

Introduction

Despite the end of the Cold War, many governments spend ever larger shares of their budgets on arms (Tan, 2010: 3). This trend could be hindered or even reversed through stricter rules and regulations (Bauer, 2010: 306). Yet, due to a ‘multilateralization’ of security after the Cold War (Krause, 1998: 1), the successful adoption and implementation of arms control agreements today is conditional on the commitment of many more states than before. This means that both perspectives and positions have diversified (Gallagher, 1997; Sands, 1997: 130) and we need an updated, systematic assessment of what determines whether states embrace or reject arms control.

In this study, I examine the effect of external threats on state support for arms control. Arms control scholars have denoted external threats, that is, threats by other states, as the main explanatory factor driving opposition to arms control (see e.g. Rosert, 2011: 257; Sagan, 1996: 54). Threatened states have an increased demand for armaments and thus reject international agreements that restrict their weapons acquisition (Glaser, 1994; Jervis, 1978; Sagan, 1996). In its most ‘hawkish’ form, this relationship has been labelled the arms control paradox: Arms control is irrelevant when it is possible, and impossible when it is needed – because states will oppose arms control when they are in dispute with each other (Gray, 1992; see also Kydd, 2000).

Due to a lack of data for more thorough cross-country analyses, contemporary empirical research is mostly limited to case studies of the Middle East (Jones, 1998; Steinberg, 1994; 2005). Extant quantitative analyses focus on actual arms acquisition rather than position-taking on arms control (e.g. Collier & Hoeffler, 2007; Jo & Gartzke, 2007; Singh & Way, 2004) and therefore only tell us little about the prospects of cooperative measures between states: while armaments and arms control are certainly linked, they are not equivalent. If an external threat leads to arms acquisition, it does not necessarily follow that it also causes opposition to arms control.

To examine the relationship between external threats and support for arms control quantitatively, I introduce a novel measure of the latter variable, combining United Nations General Assembly (UNGA) voting data with manual coding of 1,178 resolutions introduced between the 49th and 71st sessions (1994/1995 to 2016/2017). First, I assess on a five-point coding scheme whether a resolution strengthens or weakens arms control and whether divergences in voting behaviour might be driven by other conflict dimensions than the one under consideration. Afterwards, every vote by a state is assigned a position depending on the respective resolution’s categorization. This approach has the advantage that it constitutes a variance-based and more generalizable theory test than the aforementioned case studies that goes beyond a specific state or region with distinctive characteristics.

Across a variety of model specifications, I find no significant effect of external threats on arms control support. I argue that this might reflect two opposing mechanisms that cancel out each other in the event of external threats: external threats increase not only the costs, but also the benefits of arms control, as arms limitations restrict both sides of a rivalry or a dispute. While involvement in an interstate rivalry or dispute still increases the adversaries’ demand for armament, this does not translate into less support for arms control.

An alternative interpretation of the empirical results is that external threats simply do not matter for states’ position-taking on arms control. For instance, states might not perceive interstate disputes and rivalries as severe enough to alter their foreign policy preferences or they might view arms restrictions in the form of international agreements as ineffective and thus irrelevant. Either way, the findings of this study challenge realist accounts of a strictly negative effect and thus shed new light on the relationship between external security threats and arms control. Moreover, they indicate that one cannot simply derive states’ stance toward arms control from their armament decisions, which illustrates the added value of my measure.

In the second section, I first review the extant literature on the effect of external threats on position-taking on arms control and expound why the study of that relationship requires a new quantitative approach. In the third section, I describe my data, including the coding procedure for the dependent variable, as well as the research design. In the fourth section, I present and discuss the empirical results. The fifth section reflects on the implications of my empirical results and points toward directions for future research.

The effect of external threats on state support for arms control

As arms control is a contested concept that has been used in many different ways, it is necessary first to clarify what I refer to in this study. I follow Goldblat (2002: 3) in defining arms control as measures that are intended to:

(a) freeze, limit, reduce or abolish certain categories of weapons; (b) ban the testing of certain weapons; (c) prevent certain military activities; (d) regulate the deployment of armed forces; (e) proscribe transfers of some militarily important items; (f) reduce the risk of accidental war; (g) constrain or prohibit the use of certain weapons or methods of war; and (h) build up confidence among states through greater openness in military matters.

Arms control as defined above is thus understood quite broadly and also encompasses measures of disarmament and non-proliferation, for instance, the Chemical Weapons Convention and the Arms Trade Treaty. Although some arms control studies rely on a narrower definition that is sharply distinguished from disarmament, I argue that their arguments are equally applicable to the definition employed here.¹

The paramount role of external threats in arms control theory is mostly derived from (structural) realism (Glaser, 1994; Sagan, 1996). The basic assumption here is that in an anarchical international system, states are primarily driven by their need for security. With regard to their armaments policy, they mainly face two options: cooperation, that is, arms control; or defection, that is, arms build-up (see e.g. Glaser, 1994; Jervis, 1978; Kydd, 2000). This decision mainly follows from their threat perceptions: if another state is perceived as hostile, they will prefer defection to cooperation, arm themselves, and reject arms control measures (Glaser, 1994; Jervis, 1978; Sagan, 1996). This is due to two reasons: first, as states are security-seeking, the threat imposed by another state causes the need for self-defence; and second, the distrust in the adversary state will impose fear of being cheated on in potential agreements, which leads to the rejection of arms control.

This argument provides the basis for two related concepts: the security dilemma; and the aforementioned arms control paradox. The security dilemma postulates that states that feel threatened by each other might engage in a costly arms race, although it would be more beneficial for both of them to cooperate (Jervis, 1978). In a similar vein, Gray (1992) argues that arms control only has added value – that is, it could decrease armaments and prevent arms races – when states threaten each other. Yet, when threatened, states will oppose arms control, making it impossible whenever it is needed.

In line with this rationale, Steinberg (1994; 2005) traces the failure of arms control negotiations in the Middle East back to the unstable security environment and the states’ hostility toward each other. For instance, Steinberg argues that Israel’s nuclear armament and refusal to join the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) is, at least in parts, a result of its need to deter its enemies in the region (Steinberg, 1994). Similarly, Jones (1998) attributes Iran’s negative stance toward arms control to its threat perceptions and the North Korean regime justified its pursuit of nuclear weapons and NPT withdrawal with the United States nuclear threat (Wunderlich et al., 2013: 276).

Contrary to this, Glaser (1994) argues that even from a realist point of view, it should be highly context-specific whether states prefer an arms build-up to cooperation when faced with external threats. Security-maximizing states might respond to threats by refraining from armaments and seeking cooperation with their adversaries in certain situations in order to avoid a costly and risky arms race that they could lose (Glaser, 1994). If the benefits of restricting opponents’ armaments outweigh the costs of imposing these restrictions on oneself, it is perfectly rational to respond to a threat by embracing arms control.

A number of quantitative studies have examined the first part of the causal mechanism, that is, the effect of external threats on armaments. While some of them provide ambiguous results (Dunne & Perlo-Freeman, 2003; Jo & Gartzke, 2007), most identify a positive relationship between external threats and different variables related to arms acquisition. These include, inter alia, military spending (Collier & Hoeffler, 2007), arms imports (Blomberg & Tocoian, 2016), and nuclear proliferation (Fuhrmann & Horowitz, 2015; Singh & Way, 2004). Although these studies do not claim to provide any evidence for the relationship between external threats and support for arms control, they appear to speak against Glaser’s (1994) argument at first glance.

This is due to the fact that Glaser’s study as well as previous studies with a realist perspective (e.g. Jervis, 1978; Kydd, 2000) equate the decision to build up arms with defection and opposition to arms control. I argue that this is not necessarily the case. A state’s decision to cooperate or defect is certainly linked to its decision to acquire weapons or not. Nevertheless, states might reject arms control while not pursuing armaments, and, more importantly, they can build up arms and still support arms control. Kreps, Saunders & Schultz (2018) argue that, in certain situations, governments might even decide to strengthen their military capabilities precisely because they support arms control.²

The latter argument suggests that armament and arms control support are not mutually exclusive. Empirical evidence for a positive relationship between external threats and armaments does not necessarily entail a negative effect of external threats on support for arms control. To test this relationship, one thus needs to identify arms control support through other means.

Data and method

Dependent variable

As outlined above, it is not possible to derive states’ positions toward arms control from their levels of armaments. I have therefore developed an alternative approach, which relies on UNGA voting data instead to measure support for arms control. First, with the help of a student assistant, I manually code all UNGA resolutions that were drafted in the First Committee³ as well as disarmament resolutions in the plenary⁴ and adopted between the 49th and 71st UNGA sessions (1994/1995 to 2016/2017). This leads to a set of 1,178 resolutions. By means of the manual coding, which I will further describe below, I then derive states’ positions from their votes on these resolutions.⁵

This approach has several advantages in comparison to the small set of previous studies that have quantitatively measured positions on arms control. Knopf (1998) measures preferences for arms control through participation in negotiations, but his analysis focuses only on the Cold War period, the United States and its bilateral talks with the Soviet Union. Thus, it covers neither multilateral arms control nor variation across countries.

Other studies have analysed arms control treaty signature or ratification (Brender, 2018; Vaynman, 2014), arguing that this reflects states’ commitment to arms control (Brender, 2018: 16). Although treaty ratification indeed depends on policy preferences and positions to a certain extent, it is also determined by a number of other factors, such as institutional constraints (e.g. Hug & König, 2002; Kreps, Saunders & Schultz, 2018). Moreover, it only measures support for existing treaties and thus suffers from severe selection bias.

Efrat (2010) conducted a survey among officials from 118 countries to assess the respective governments’ preferences on controlling the trade of small arms and light weapons. While this is undoubtedly a useful measure, the survey was only implemented once, and does not capture variation over time. In addition, it focuses on a very specific issue and not support for arms control in general.

Instead, I measure arms control support through UNGA voting behaviour, which has become the standard data source for measuring foreign policy positions (Bailey, Strezhnev & Voeten, 2017: 430). As the UNGA plays a significant role in international arms control negotiations and covers the entire spectrum of arms control issues (Müller, Below & Wisotzki, 2013; Thakur, 2017), it constitutes a useful forum for studying positions on arms control. In contrast to speeches or co-sponsorships, voting allows for the comparison of all states’ positions on the same issues (Finke, 2021: 6–7). This data source seems to better reflect the overall topic of arms control and states’ preferences thereon.⁶

That said, most existing methods for deriving foreign policy preferences from UNGA voting merely measure the states’ proximity to each other, not their stance toward the subject under consideration. They indicate, for instance, whether a government votes more with the states of the Global North (e.g. Kim & Russett, 1996), the Western states (e.g. Voeten, 2000), or specific countries, particularly the United States (e.g. Carter & Stone, 2014; Dreher & Jensen, 2013) and China (e.g. Carmody, Dasandi & Mikhaylov, 2019; Flores-Macías & Kreps, 2013). Accordingly, the analysis of support for arms control requires a different measurement technique.

This poses two challenges. First, not every resolution unrestrictedly strengthens arms control. For instance, resolution A/RES/57/54 criticizes ‘the growing proliferation of ad hoc and exclusive export control regimes and arrangements for dual-use goods and technologies’ and demands that the transfer of dual-use goods be less restricted (United Nations General Assembly, 2002: 1). Hence, states that are in favour of arms control might sometimes vote against resolutions.

To solve a similar problem in their study on human rights resolutions, Boockmann & Dreher (2011) use the votes of four states that are known to be very supportive of human rights as a benchmark for the ‘pro human rights’ vote. In other words, they define the voting option that reflects one extreme position on their conflict dimension, and assess other states’ positions through their agreement with this benchmark.⁷ Yet, their

Figure 1.

Coding procedure

approach requires the existence of – and knowledge about – one or more countries that always vote for one of the two extreme positions.

Moreover, it needs all disagreements to reflect the conflict dimension under consideration. Divergences in votes on arms control resolutions might not only originate from dissent over support or non-support for arms control. This constitutes the second challenge. For instance, states also disagree on which issues should be given priority: non-proliferation or disarmament (Barnum & Lo, 2020), multilateral or bilateral agreements (Krause, 1998: 17), nuclear or conventional arms control (Meyer, 2016), vertical or horizontal non-proliferation (Schörnig, 2017: 966), and so on.

To address these challenges, all 1,178 resolutions are manually coded. I assess on a five-point manual coding scheme whether the resolution strengthens or weakens arms control, and whether divergences in votes might result from other conflict dimensions than the one I aim to measure. The coding procedure can be regarded as a decision tree with up to five steps (see Figure 1; for a detailed description of the coding procedure, see Online appendix A).

Inspired by Boockmann & Dreher (2011), each vote is assigned a position depending on the category to which the respective resolution belongs (see Table I; see Figure 1 for examples).⁸ Category 1 contains resolutions that clearly strengthen arms control. States that support arms control will vote in favour. The resolutions in Category 2 tend to strengthen arms control, but include passages that some states might consider counter-productive and make them abstain – despite being clearly supportive of arms control. Categories 3 and 4 contain those resolutions that weaken arms control. Category 4 clearly does so, whereas Category 3 also includes passages that can be regarded as supportive of

Table I.

Category	Yes	Abstain	No
1	In favour (1)	Neutral (0.5)	Against (0)
2	In favour (1)	In favour (1)	Neutral (0.5)
3	Against (0)	In favour (1)	In favour (1)
4	Against (0)	Neutral (0.5)	In favour (1)
0	–	–	–

Independent variable

As my first measure of external threats, I use a binary variable indicating whether a state was involved in a militarized interstate dispute (MID) (Palmer et al., 2020) in a given year or not. The MID data includes militarized conflicts between states ranging from threats to use force to full-scale interstate war. For my second measure of external threats, I rely on the peace data (PD) (Diehl, Goertz & Gallegos, 2021; Goertz, Diehl & Balas, 2016). This dataset codes interstate rivalries through a number of different indicators, such as conflicting issues, a history of MIDs and the absence of diplomatic relations and communication (Diehl, Goertz & Gallegos, 2021). Again, I use a binary indicator measuring whether a state was involved in a rivalry in a given year or not. Previous studies have used both variables to measure external threats (e.g. Arbatli & Arbatli, 2014; Thies, 2007).

The first variable captures actual behaviour, including minor incidents and short-term hostilities, whereas the second variable identifies latent rivalries, that is, long-lasting hostile relationships that do not necessitate militarized action every year. Comparing both measures yields a moderate correlation of around 0.47, which indicates that there is some overlap, but they do differ to a substantial degree. By using both data sources, I capture different types of threats that might induce different kinds of reactions and hence vary in their effect on arms control support.

Empirical model

Studies of UNGA voting behaviour have pursued different approaches, treating the dependent variable as continuous, binary,¹² or ordinal with either the country–year or the single vote as the unit of analysis (e.g. Boockmann & Dreher, 2011; Dreher & Jensen, 2013; Wang, 1999). My main model is on the vote level, as this allows me to include resolution-specific controls, and linear, due to its simplicity and more straightforward interpretability. Yet, ordered and binary logit models on the vote level as well as a linear model on the country–year level are included as robustness checks and yield the same results as the main analysis.

The data are clustered on country, year and (repeated) resolution levels.¹³ Therefore, I include random effects on these three levels and thus run a linear mixed-effects model. The independent variables, especially interstate rivalries, mostly vary across rather than within countries. Hence, fixed effects would largely cancel out meaningful variation. Nevertheless, a model with fixed effects for country and year is included as a robustness check and reproduces the findings from the random-effects model.¹⁴

To avoid omitted variable bias,¹⁵ I introduce a number of control variables that could affect the likelihood of external threats and which are prevalent in the literature on arms control and armaments. On the country–year level, I control for the level of electoral democracy (Coppedge et al., 2021), gross domestic product per capita (United Nations Statistics Division, 2020), trade openness (Feenstra, Inklaar & Timmer, 2015), government ideology (Cruz, Keefer & Scartascini, 2021), national material capabilities (CINC) (Singer, 1988; Singer, Bremer & Stuckey, 1972), and intrastate conflict (Gleditsch et al., 2002; Pettersson et al., 2021).

As the main model not only incorporates variation over time but also across countries, I further include a number of (mostly) time-constant controls. These include region (Gleditsch et al., 2002; Pettersson et al., 2021), European Union membership and North Atlantic Treaty Organization (NATO) membership, and nuclear weapons possession.

To account for agenda changes and enhance statistical efficiency, I also include four variables on the resolution level: topic, category, and global relevance, all of them assigned in the coding procedure, as well as salience, measured through the number of UNGA speeches mentioning a resolution (Finke, forthcoming).¹⁶

Empirical analysis

As a first step, I run bivariate analyses with both threat measures (see Online appendix F for all regression tables, including a stepwise introduction of control variables). They show a negative and highly significant relationship between interstate disputes as well as rivalries and state support for arms control. However, this relationship disappears when including random effects and control variables. In the main model (see Figures 2 and 3), the dummies for interstate disputes and rivalries yield extremely small coefficients that remain far from significant (p ≈ 0.74; p ≈ 0.52). This indicates that the negative relationship between external threats and arms control support in the bivariate model is induced by omitted variable bias. In other words, I do not find any effect of external threats on support for arms control, in the form of neither actual militarized action nor latent rivalries. Thus, the analysis contradicts the realist notion that external threats are the main factor leading states to oppose arms control.

Following Rainey (2014), I will not only discuss the results’ significance levels but also their confidence intervals to assess whether this is a true null finding or whether the results are consistent with meaningful effects. First, I define an effect size of -0.005 as meaningful, which corresponds to a negative vote shift in one out of 100 resolutions. While this might not appear to be a substantial decrease, one should keep in mind that more than half of the resolutions are adopted by consensus, and all resolutions are supported by large majorities. Accordingly, such an effect size would, on average, move a country between five and six positions downwards in the country level ranking.

Then, I examine whether the independent variables’ confidence intervals include the value of -0.005. This is neither the case for the MID dummy (-0.002; 0.002) nor for the rivalry variable (-0.004; 0.002). Thus, not

Figure 2.

Main analysis – militarized interstate dispute data (showing country and country–year level variables only)

Figure 3.

Main analysis – peace data (showing country and country–year level variables only)

only do I find no evidence for a significant negative relationship between external threats and arms control support, but the results also speak in favour of a true null finding – or ‘a negligible effect’ (Rainey, 2014: 1083) – that is not merely driven by a lack of statistical power.

The null findings remain robust across a large variety of robustness checks. First, as indicated above, nuclear weapons possession and CINC score are related to armaments and could thus not only confound, but also mediate the relationship between external threats and arms control support. Excluding them from the analysis also yields insignificant results. Thus, the null findings are not caused by the introduction of control variables that could potentially suppress the mechanism connecting external threats and opposition to arms control.

Next, instead of binary variables, I rely on two alternative threat measures derived from the MID and PD. First, both datasets contain ordinal measures that indicate different levels of threat severity.¹⁷ Second, I use count variables identifying the number of threats a country experiences in a given year. Moreover, I replace the independent variables with measures from other data sources. These include the continuous interstate hostility measure introduced by Terechshenko (2020) as well as binary indicators of international rivalries identified by Thompson & Dreyer (2011) and of interstate conflicts (Gleditsch et al., 2002; Pettersson et al., 2021). None of the alternative measures of external threats produce significant results. This indicates that the effect, or rather the lack thereof, holds irrespective of the operationalization of external threats.

The same holds true when running ordered logit or binary logit instead of linear models and when analysing the data on the country–year level. Using country-fixed and year-fixed instead of random effects also leads to insignificant results. So does lagging the independent and control variables – except for those on the resolution level – by one year and removing consensus votes from the analysis.

Furthermore, I run additional regression models that include only resolutions in the first category or the first two categories, respectively. Thus, I assess whether external threats only have an effect on voting with regard to those resolutions that actually strengthen arms control. Both analyses reproduce the insignificant results. This is also the case with regard to analyses that exclude nuclear weapons possessors and NATO members, respectively (i.e. states that cast a large proportion of the negative votes).

Finally, I run separate regression models for nuclear and non-nuclear votes. Although scholars have argued that the negative effect of external threats on arms control support should hold regardless of the type of weapons under consideration (Gray, 1992), most realist analyses focus on nuclear arms control (e.g. Glaser, 1994; Jervis, 1978; Kydd, 2000). Interestingly, this yields results that are less clear-cut. Both independent variables remain insignificant in the non-nuclear model, and there is no significant association between interstate rivalries and support for nuclear arms control. Yet, the MID dummy is significant and positive in the analysis of nuclear votes.

A closer look at the data indicates, however, that one should not interpret this association in causal terms. Some states are involved in an MID in several isolated years with a large number of positive votes on nuclear arms control, while voting more negatively in the preceding and following years. Lagging the independent variables in this model by one year supports this caveat, as it does not reproduce the significant result. I therefore refrain from interpreting this finding in depth. Nevertheless, it certainly speaks against a negative relationship between external threats and arms control support. Moreover, it might indicate that states do not behave the same regardless of the type of arms control under consideration, providing a promising starting point for future studies.

While not all model specifications definitely rule out a meaningful effect (see Online appendix F), the large majority, including the main models, do. More importantly though, none of them yield a significant negative association between external threats and support for arms control. In sum, the robustness checks thus confirm the findings from the main model and contradict the realist argument that the involvement in interstate disputes and rivalries decreases states’ arms control support.

There could be several reasons for the lack of a significant negative relationship. External threats might simply not affect support for arms control at all. Arms control in its current form has been criticized repeatedly for its ineffectiveness (see e.g. Krause, 2018; Schörnig, 2017). States might not regard arms control instruments as effective constraints to their armaments and, as a consequence, do not adapt their positions when faced with external threats.¹⁸ Yet, a closer look at the control variables suggests the opposite: for instance, nuclear weapons possession and NATO membership are negatively related to arms control support. This indicates that states do consider arms control to be a relevant topic that can indeed affect their armament policies.

Another possibility is that not all interstate disputes, rivalries and conflicts constitute severe security threats. This notion runs counter to structural realist ideas: Gray (1992) claims that the arms control paradox should hold for both major powers and developing countries and for severe conflicts as well as ‘international rivalry “as usual”’ (Gray, 1992: 19). It also seems doubtful that, as previous studies have shown, such external threats raise armaments’ levels, but are not grave enough to translate into changes of arms control preferences.

I argue instead that the empirical analysis does not necessarily speak against the causal mechanism postulated above. A security threat by another state does increase the need for armaments and thus the costs of arms control. However, arms control not only restricts one state, but both states threatening each other. As Glaser (1994) has already argued, an external threat also creates a demand for restricting the adversary state’s armaments, which raises the benefits of arms control.

Even if the rival state rejects cooperative measures or is expected to do so, arms control can be a useful tool for exerting pressure and fostering cooperation through norm diffusion (Gibbons, 2018: 11–12). Moreover, many arms control measures have effective consequences for all states, even those that have not signed or ratified the respective agreement.¹⁹ Whether these benefits or the aforementioned costs of arms control then predominate in the face of an external threat depends on the context. In the aggregate, both mechanisms cancel out each other, which causes the null findings.

The results thus speak against a strictly negative relationship between the two variables. This holds regardless of whether there are indeed two opposing mechanisms at play or whether, as described above, external threats simply do not affect arms control support. Accordingly, this study speaks against the structural realist argument that external threats are the key factor explaining states’ opposition to arms control.

Moreover, if one equates armament and opposition to arms control, the null findings appear to contradict previous studies that have identified a significant effect of external threats on the former variable. I argue that the results rather indicate that arms build-up and a preference for cooperative measures are not mutually exclusive. Armaments constitute a response to security threats that, unless it violates existing agreements, does not necessarily run counter to a supportive position toward arms control. Arms build-up could even be a strategy to create incentives for the adversary state to agree to arms restrictions. Thus, the empirical results illustrate that armaments decisions and positions toward arms control should not be conflated and that it is worthwhile studying the latter variable.

Conclusion

The aim of this article was to investigate whether external threats affect state support for arms control. To assess this relationship quantitatively, I introduced a novel measure of the latter variable that combines UNGA voting data with manual coding of 1,178 UNGA resolutions. The empirical analysis has shown that neither interstate disputes nor interstate rivalries, two common measures of external threats, are significantly related to arms control support. This finding remains robust across a large variety of model specifications.

This study contributes to the academic literature on arms control in several important ways. Given that previous studies mostly identify a positive effect of external threats on weapons acquisition, the analysis illustrates that a decision to build up arms does not always induce opposition to cooperative and restrictive measures. States might reject or embrace arms control irrespective of their own level of armaments. Hence, one cannot simply derive their arms control preferences from their armaments decisions, which demonstrates the added value of the indicator introduced here.

Moreover, the null findings speak against the argument that external threats are the main explanatory factor of opposition to arms control. This is good news for proponents of arms control, as it contradicts the arms control paradox: if external threats do not inevitably decrease support for cooperative measures, then arms control might not be impossible to achieve – even when it is needed.

Yet, one needs to consider a couple of limitations. First, this study only examines support for and not advocacy or salience of arms control. The importance that states attach to arms control, and how it is affected by external threats, is something that future studies could address – for instance relying on UNGA speeches or co-sponsorships (Finke, 2021, forthcoming). Still, although I do not measure the importance attached to arms control, I illustrate that external threats do not change states’ substantive positions on this topic. This constitutes an important first step in the study of their policy agendas toward arms control.

Second, the null findings could point toward two different theoretical explanations. They might indicate that external threats do not affect arms control support, for instance because states regard arms control as ineffective in restricting their own armaments. However, the results could also reflect an effect that is not strictly negative, but context-specific. An external threat not only raises the costs, but also the benefits of arms control, as restrictions on armaments do not only restrain the threatened state, but both sides of a dispute or rivalry.

Future research should further examine the circumstances under which external threats affect support for arms control and in which ways. Analyses should, inter alia, take into account characteristics of the adversary states or the specific form of arms control under consideration. This will shed further light on whether the empirical results are driven by two contrasting mechanisms or whether external threats and arms control support are in fact not related at all. Either way, this study contradicts the notion of a strictly negative relationship between external threats and state support for arms control. This implies that external threats alone do not explain states’ positions on arms control.

Returning to the initial puzzle of what determines state support for arms control, this begs the question of which factors play a role instead. While the analysis suggests that armaments and arms control support are not equivalent, regression results as well as country level scores indicate that the two variables are related. Beyond this, states’ alliances, for instance with NWS, can also shape their views on the necessity of controlling these weapons. Thus, membership in international organizations, such as NATO, might also affect states’ position-taking on arms control.

Moving away from structural realist ideas, which largely ignore the domestic arena, the data suggest that one should also take into account developments within states: democratic states seem to be more favourable toward arms control. Moreover, it could also be worthwhile investigating internal threats, given that most arms are used in intrastate instead of interstate conflicts. Future studies should thus investigate these as well as other potential determinants. In this regard, the position measurement introduced in this study provides a useful tool.

Footnotes

Replication data

The dataset and code for the empirical analysis in this article, along with the online appendix, are available at . All analyses were conducted using R.

Acknowledgements

I thank Ida Braad Albek for valuable research assistance. I also thank Laurits Florang Aarslew, Troy Saghaug Broderstad, Valentin Daur, Daniel Finke, Kristian Vrede Skaaning Frederiksen, Lasse Egendal Leipziger, Casper Sakstrup, Svend-Erik Skaaning, Jakob Tolstrup, the Comparative Politics Section at Aarhus University’s Department of Political Science, panel members at the Jan Tinbergen Peace Science Conference and the European Consortium for Political Research General Conference as well as two anonymous reviewers for their excellent comments and suggestions.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Department of Political Science, Aarhus University.

ORCID iD

Tobias Risse

Notes

References

Arbatli

Cemal Eren

Arbatli

Ekim

(2014) External threats and political survival: Can dispute involvement deter coup attempts? Conflict Management and Peace Science 33(2): 115–152.

Bailey

Michael A

Strezhnev

Anton

Voeten

Erik

(2017) Estimating dynamic state preferences from United Nations voting data. Journal of Conflict Resolution 61(2): 430–456.

Barnum

Miriam

James

(2020) Is the NPT unraveling? Evidence from text analysis of review conference statements. Journal of Peace Research 57(6): 740–751.

Bauer

Sibylle

(2010) Post-Cold War control of conventional arms. In: Tan

Andrew TH

(ed.) The Global Arms Trade: A Handbook. Abingdon: Routledge, 306–318.

Becker-Jakob

Una

Hofmann

Gregor

Müller

Harald

Wunderlich

Carmen

(2013) Good International Citizens. Canada, Germany, and Sweden: Müller & Wunderlich, 207–245.

Becker

Una

Müller

Harald

Wisotzki

Simone

(2008) Democracy and nuclear arms control—destiny or ambiguity? Security Studies 17(4): 810–854.

Blomberg

Brock

Tocoian

Oana

(2016) Who gets the guns? How conflicts and politics drive international arms transfers. Claremont McKenna College Robert Day School of Economics and Finance Research Paper, No. 2765879 (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2765879).

Boehmer

Charles R

Sobek

David

(2005) Violent adolescence: State development and the propensity for militarized interstate conflict. Journal of Peace Research 42(1): 5–26.

Boockmann

Bernhard

Dreher

Axel

(2011) Do human rights offenders oppose human rights resolutions in the United Nations? Public Choice 146(3–4): 443–467.

10.

Boryczka

Jocelyn M

Mohanan

Pearson

Frederic S

Weigand

Jeffrey D

(1997) Cultural and strategic factors in South Asian nuclear arms control. Journal of Political and Military Sociology 25(2): 279–303.

11.

Brazys

Samuel

Panke

Diana

(2017) Analysing voting inconsistency in the United Nations General Assembly. Diplomacy & Statecraft 28(3): 538–560.

12.

Brender

Agnes

(2018) Determinants of international arms control ratification. ILE Working Paper Series, No. 17 (https://www.econstor.eu/handle/10419/183594).

13.

Carmody

Pádraig

Dasandi

Niheer

Mikhaylov

Slava Jankin

(2019) Power plays and balancing acts: The paradoxical effects of Chinese trade on African foreign policy positions. Political Studies 68(1): 224–246.

14.

Carter

David B

Stone

Randall W

(2014) Democracy and multilateralism: The case of vote buying in the UN General Assembly. International Organization 69(1): 1–33.

15.

Collier

Paul

Hoeffler

Anke

(2007) Unintended consequences: Does aid promote arms races? Oxford Bulletin of Economics and Statistics 69(1): 1–27.

16.

Coppedge

Michael

Gerring

John

Knutsen

Carl Henrik

Lindberg

Staffan I

Teorell

Jan

Alizada

Nazifa

Altman

David

, et al. (2021) V-Dem country–year dataset v11 (https://www.v-dem.net/en/data/archive/previous-data/v-dem-dataset/).

17.

Crail

Peter

(2011) IAEA sends Syria nuclear case to UN. Arms Control Today 41(6): 22–24.

18.

Cruz

Cesi

Keefer

Philip

Scartascini

Carlos

(2021) Database of Political Institutions 2020 . Inter-American Development Bank Research Department (https://publications.iadb.org/en/database-political-institutions-2020-dpi2020).

19.

D’Acquisto

Germana

(2017) A Linguistic Analysis of Diplomatic Discourse: UN Resolutions on the Question of Palestine. Newcastle upon Tyne: Cambridge Scholars Publishing.

20.

Diehl

Paul F

Goertz

Gary

Gallegos

Yahve

(2021) Peace data: Concept, measurement, patterns, and research agenda. Conflict Management and Peace Science 38(5): 605–624.

21.

Dreher

Axel

Jensen

Nathan M

(2013) Country or leader? Political change and UN General Assembly voting. European Journal of Political Economy 29: 183–196.

22.

Dunne

Paul

Perlo-Freeman

Sam

(2003) The demand for military spending in developing countries. International Review of Applied Economics 17(1): 23–48.

23.

Efrat

Asif

(2010) Toward internationally regulated goods: Controlling the trade in small arms and light weapons. International Organization 64(1): 97–131.

24.

Feenstra

Robert C

Inklaar

Robert

Timmer

Marcel P

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

25.

Fey

Marco

Hellmann

Andrea

Klinke

Friederike

Plümmer

Franziska

Rauch

Carsten

(2013) Established and Rising Great Powers: The United States . Russia, China, and India: Müller & Wunderlich, 163–206.

26.

Finke

Daniel

(2021) Regime type and co-sponsorship in the UN General Assembly. International Interactions 47(3): 559–578.

27.

Finke

Daniel

(forthcoming) Party ideologies and UN debates. Party Politics. DOI: 10.1177/13540688221074494.

28.

Flores-Macías

Gustavo A

Kreps

Sarah E

(2013) The foreign policy consequences of trade: China’s commercial relations with Africa and Latin America, 1992–2006. The Journal of Politics 75(2): 357–371.

29.

Fuhrmann

Matthew

Horowitz

Michael C

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

30.

Gallagher

Nancy W

(1997) Bridging the gaps on arms control. Contemporary Security Policy 18(2): 1–24.

31.

Gibbons

Rebecca Davis

(2018) The humanitarian turn in nuclear disarmament and the Treaty on the Prohibition of Nuclear Weapons. Nonproliferation Review 25(1–2): 11–36.

32.

Glaser

Charles L

(1994) Realists as optimists: Cooperation as self-help. International Security 19(3): 50–90.

33.

Gleditsch

Kristian Skrede

Salehyan

Idean

Schultz

Kenneth

(2008) Fighting at home, fighting abroad: How civil wars lead to international disputes. Journal of Conflict Resolution 52(4): 479–506.

34.

Gleditsch

Nils Petter

Wallensteen

Peter

Eriksson

Mikael

Sollenberg

Margareta

Strand

Håvard

(2002) Armed conflict 1946–2001: A new dataset. Journal of Peace Research 39(5): 615–637.

35.

Goertz

Gary

Diehl

Paul F

Balas

Alexandru

(2016) The Puzzle of Peace: Explaining the Rise of Peace in the International System. New York: Oxford University Press.

36.

Goldblat

Jozef

(2002) Arms Control: The New Guide to Negotiations and Agreements. London: SAGE.

37.

Gray

Colin S

(1992) House of Cards: Why Arms Control Must Fail. Ithaca, NY: Cornell University Press.

38.

Green

Michael J

Furukawa

Katsuhisa

(2000) New ambitions, old obstacles: Japan and its search for an arms control strategy. Arms Control Today 30(6): 17–24.

39.

Götz

Elias

MacFarlane

Neil

(2018) Russia’s role in world politics: Power, ideas, and domestic influences. International Politics 56(6): 713–725.

40.

Hug

Simon

König

Thomas

(2002) In view of ratification: Governmental preferences and domestic constraints at the Amsterdam intergovernmental conference. International Organization 56(2): 447–476.

41.

Häge

Frank

Hug

Simon

(2016) Consensus decisions and similarity measures in international organizations. International Interactions 42(3): 503–529.

42.

Jervis

Robert

(1978) Cooperation under the security dilemma. World Politics 30(2): 167–214.

43.

Dong-Joon

Gartzke

Erik

(2007) Determinants of nuclear proliferation. Journal of Conflict Resolution 51(1): 167–194.

44.

Jones

Peter

(1998) Iran’s threat perceptions and arms control policies. Nonproliferation Review 6(1): 39–55.

45.

Kim

Soo Yeon

Russett

Bruce

(1996) The new politics of voting alignments in the United Nations General Assembly. International Organization 50(4): 629–652.

46.

Knopf

Jeffrey W

(1998) Domestic sources of preferences for arms cooperation: The impact of protest. Journal of Peace Research 35(6): 677–695.

47.

Krause

Keith

(1998) Cross-cultural dimensions of multilateral non-proliferation and arms control dialogues: An overview. Contemporary Security Policy 19(1): 1–22.

48.

Krause

Keith

(2018) Arms control and disarmament. In: Weiss

Thomas G

Daws

Sam

(eds) The Oxford Handbook on the United Nations. Oxford: Oxford University Press, 383–395.

49.

Krause

Keith

Latham

Andrew

(1998) Constructing nonproliferation and arms control: The norms of Western practice. Contemporary Security Policy 19(1): 23–54.

50.

Kreps

Sarah E

Saunders

Elizabeth N

Schultz

Kenneth A

(2018) The ratification premium: Hawks, doves, and arms control. World Politics 70(4): 479–514.

51.

Kydd

Andrew

(2000) Arms races and arms control: Modeling the hawk perspective. American Journal of Political Science 44(2): 228–244.

52.

Mattes

Michaela

Leeds

Brett Ashley

Carroll

Royce

(2015) Leadership turnover and foreign policy change: Societal interests, domestic institutions, and voting in the United Nations. International Studies Quarterly 59(2): 280–290.

53.

Meyer

Paul

(2016) Hard and soft linkages between nuclear and conventional disarmament. In: United Nations Office for Disarmament Affairs (ed.) UNODA Occasional Papers No.28: Rethinking General and Complete Disarmament in the Twenty-First Century. New York: United Nations.

54.

Müller

Harald

Wunderlich

Carmen

, eds (2013) Norm Dynamics in Multilateral Arms Control: Interests, Conflicts, and Justice. Athens, GA: University of Georgia Press.

55.

Müller

Harald

Below

Alexis

Wisotzki

Simone

(2013) Beyond the state: Nongovernmental organizations, the European Union, and the United Nations, in Müller & Wunderlich, 296–336.

56.

Nystuen

Gro

Egeland

Kjølv

(2019) The potential of the Arms Trade Treaty to reduce violations of international humanitarian law and human rights law. In: Bailliet

Cecilia M

(ed.) Research Handbook on International Law and Peace. Cheltenham: Edward Elgar, 267–283.

57.

Palmer

Glenn

McManus

Roseanne W

D’Orazio

Vito

Kenwick

Michael R

Karstens

Mikaela

Bloch

Chase

Dietrich

Nick

Kahn

Kayla

Ritter

Kellan

Soules

Michael J

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

58.

Panke

Diana

(2014) The UNGA—a talking shop? Exploring rationales for the repetition of resolutions in subsequent negotiations. Cambridge Review of International Affairs 27(3): 442–458.

59.

Pettersson

Therése

Davies

Shawn

Deniz

Amber

Engström

Garoun

Hawach

Nanar

Högbladh

Stina

Öberg

Margareta Sollenberg Magnus

(2021) Organized violence 1989–2020, with a special emphasis on Syria. Journal of Peace Research 58(4): 809–825.

60.

Rainey

Carlisle

(2014) Arguing for a negligible effect. American Journal of Political Science 58(4): 1083–1091.

61.

Rosert

Elvira

(2011) Rüstung, Rüstungskontrolle und Abrüstung [Armament, arms control, and disarmament]. In: Schlotter

Peter

Wisotzki

Simone

(eds) Friedens- und Konfliktforschung (Peace and Conflict Research) . Baden-Baden: Nomos, 252–281. [In German.]

62.

Sagan

Scott D

(1996) Why do states build nuclear weapons? Three models in search of a bomb. International Security 21(3): 54–86.

63.

Sakaki

Alexandra

Maslow

Sebastian

(2020) Japan’s new arms export policies: Strategic aspirations and domestic constraints. Australian Journal of International Affairs 74(6): 649–669.

64.

Sands

Amy

(1997) The impact of governmental context on negotiation and implementation: Constraints and opportunities for change. Contemporary Security Policy 18(2): 116–137.

65.

Schörnig

Niklas

(2017) Rüstung, Rüstungskontrolle und internationale Politik. In: Sauer

Frank

Masala

Carlo

(eds) Handbuch Internationale Beziehungen (Handbook of International Relations) . Wiesbaden: Springer, 959–990. [In German.]

66.

Singer

J David

(1988) Reconstructing the Correlates of War dataset on material capabilities of states, 1816–1985. International Interactions 14(2): 115–132.

67.

Singer

J David

Bremer

Stuart

Stuckey

John

(1972) Capability distribution, uncertainty, and major power war, 1820–1965. In: Russett

Bruce

(ed.) Peace, War, and Numbers. Beverly Hills: SAGE, 19–48.

68.

Singh

Sonali

Way

Christopher R

(2004) The correlates of nuclear proliferation: A quantitative test. Journal of Conflict Resolution 48(6): 859–885.

69.

Steinberg

Gerald M

(1994) Israeli arms control policy: Cautious realism. Journal of Strategic Studies 17(2): 1–16.

70.

Steinberg

Gerald M

(2005) Realism, politics and culture in Middle East arms control negotiations. International Negotiation 10(3): 487–512.

71.

Stinnett

Douglas M

Diehl

Paul F

(2001) The path(s) to rivalry: Behavioral and structural explanations of rivalry development. Journal of Politics 63(3): 717–740.

72.

Tan

Andrew TH

(2010) The global arms trade. In: Tan

Andrew TH

(ed.) The Global Arms Trade: A Handbook. Abingdon: Routledge, 3–12.

73.

Terechshenko

Zhanna

(2020) Hot under the collar: A latent measure of interstate hostility. Journal of Peace Research 57(6): 764–776.

74.

Thakur

Ramesh

(2017) The United Nations, Peace and Security: From Collective Security to the Responsibility to Protect (2nd edition). Cambridge: Cambridge University Press.

75.

Thies

Cameron G

(2007) The political economy of state building in Sub-Saharan Africa. The Journal of Politics 69(3): 716–731.

76.

Thompson

William R

Dreyer

David

(2011) Handbook of International Rivalries. Washington, DC: CQ Press.

77.

Trapp

(2014) Elimination of the chemical weapons stockpile of Syria. Journal of Conflict and Security Law 19(1): 7–23.

78.

United Nations General Assembly (1978) Final document of the tenth special session of the General Assembly (A/RES/S-10/2, https://digitallibrary.un.org/record/218448?ln=en).

79.

United Nations General Assembly (1994a) Prevention of an arms race in outer space (A/RES/49/74, https://digitallibrary.un.org/record/166831?ln=en).

80.

United Nations General Assembly (1994b) Rationalization of the work and reform of the agenda of the First Committee (A/RES/49/85, https://digitallibrary.un.org/record/166807?ln=en).

81.

United Nations General Assembly (1994c) The risk of nuclear proliferation in the Middle East (A/RES/49/78, https://digitallibrary.un.org/record/166805?ln=en).

82.

United Nations General Assembly (1995) Observance of environmental norms in the drafting and implementation of agreements on disarmament and arms control (A/RES/50/70 M, https://digitallibrary.un.org/record/239915?ln=en).

83.

United Nations General Assembly (1996) The risk of nuclear proliferation in the Middle East (A/RES/51/48, https://digitallibrary.un.org/record/230524?ln=en).

84.

United Nations General Assembly (1998) Transparency in armaments (A/RES/53/77 S, https://digitallibrary.un.org/record/265396?ln=en).

85.

United Nations General Assembly (2002) Role of science and technology in the context of international security and disarmament (A/RES/57/54, https://digitallibrary.un.org/record/479476?ln=en).

86.

United Nations General Assembly (2016a) Nuclear disarmament (A/RES/71/63, (https://digitallibrary.un.org/record/852993?ln=en).

87.

United Nations General Assembly (2016b) Promotion of multilateralism in the area of disarmament and non-proliferation (A/RES/71/61, https://digitallibrary.un.org/record/852992?ln=en).

88.

United Nations (2018) The Committees: Disarmament for a peaceful world, defines work of UN General Assembly's First Committee (https://news.un.org/en/story/2018/12/1029221).

89.

United Nations (2019) Seventy-fourth session of the General Assembly: First Committee: Disarmament and international security: Press releases, draft resolutions, statements, Secretary-General’s reports, and side events (https://www.un.org/disarmament/meetings/firstcommittee-74/).

90.

United Nations (2020) Special political and decolonization (Fourth Committee) (https://www.un.org/en/ga/fourth/).

91.

United Nations (2021) Resolutions of the General Assembly (https://research.un.org/en/docs/ga/resolutions).

92.

United Nations Statistics Division (2020) Per Capita GDP at constant 2015 prices in US Dollars (https://unstats.un.org/unsd/snaama/downloads).

93.

Vaynman

Jane Eugenia

(2014) Enemies in agreement: Domestic politics, uncertainty, and cooperation between adversaries. Unpublished PhD dissertation, Harvard University (http://nrs.harvard.edu/urn-3:HUL.InstRepos:13070027).

94.

Voeten

Erik

(2000) Clashes in the Assembly. International Organization 54(2): 185–215.

95.

Voeten

Erik

(2013) Data and analyses of voting in the United Nations General Assembly. In: Reinalda

Bob

(ed.) Routledge Handbook of International Organization. Abingdon: Routledge, 54–66.

96.

Wang

(1999) US. foreign aid and UN voting: An analysis of important issues. International Studies Quarterly 43(1): 199–210.

97.

Wunderlich

Carmen

Hellmann

Andrea

Müller

Daniel

Reuter

Judith

Schmidt

Hans-Joachim

(2013) Non-aligned reformers and revolutionaries: Egypt, South Africa, Iran, and North Korea, in Müller & Wunderlich, 246–295.

External threats and state support for arms control

Abstract

Keywords

Introduction

The effect of external threats on state support for arms control

Data and method

Dependent variable

Independent variable

Empirical model

Empirical analysis

Conclusion

Footnotes

Replication data

Acknowledgements

Funding

ORCID iD

Notes

References