ASSESSING THE BENEFITS AND BURDENS OF NUCLEAR LATENCY

Transcription

1 ASSESSING THE BENEFITS AND BURDENS OF NUCLEAR LATENCY Rupal Mehta Gene Gerzhoy Rachel Whitlark Annual Meeting of the Peace Science Society International November 2015

2 MOTIVATION: LATENCY AFTER THE JCPOA

3 EXISTING LITERATURE Early literature used industrial capacity proxies to capture latency (Meyer 1982, Jo and Gartzke 2007, 2009). Sagan 2010 critique: overlooking key components of latency - enrichment and reprocessing (ENR). Fuhrmann and Tkach 2015: new time-series dataset of ENR capabilities But there has been no effort yet to employ these data to test earlier theories about the affect of nuclear latency on international relations.

4 CONVENTIONAL WISDOMS Virtual Deterrence Theory Levite 2003: hedging offers benefits of virtual arsenal States coerce greater commitments from allies and concessions by threatening to acquire nuclear weapons in bargaining situations Latent Provocation Theory Narang 2014: deterrence only with specific postures Gilpin 1981, Monteiro & Debs 2014: latency creates incentives for preventive attack Nuclear Latency Dataset (ENR capacity) Fuhrmann & Tkach (2015): Reduces likelihood of being targeted in MIDs; increases likelihood of initiating MIDs.

5 RESEARCH QUESTIONS What do states gain from possessing nuclear latency? Specifically, how do latent nuclear capabilities in the form of operational enrichment and reprocessing (ENR) facilities affect: a) interstate deterrence and coercion b) bargaining outcomes short of conflict

6 OUR APPROACH Examine the competing logics: Virtual Deterrence & Latent Provocation Exploit the variation among states over time to quantitatively examine how latency impacts interstate relations. Building on Fuhrmann and Tkach: use a more strict coding scheme Requires presence of at least 1 operational facility in given stateyear to be considered latent. Fuhrmann & Tkach (2015) countries are listed as 1 on latency if they operate ENR facilities of any size in a given year, and 0 otherwise (dotted line).

7 COMPETING LOGICS: OBSERVABLE IMPLICATIONS States that acquire nuclear latency are more likely to: Receive military assistance from the United States Receive economic assistance from the United States Be the target of U.S. economic sanctions Initiate MIDs Be the target of MIDs Issue successful compellent threats

8 RESEARCH DESIGN & DATA ANALYSIS Independent Variable: Latency Operational ENR Capabilities Operationalization of Latency 0/1 Dichotomous replication with new data 0/1 Operational ENR facilities in a given state-year State is coded as latent if an ENR facility is operational in a given state-year

9 DEPENDENT VARIABLES Alliance Considerations Military Aid from the U.S. Economic Aid from the U.S. Coercive Bargaining or Punishments MIDs initiation Target of MIDs Success in Issuance of Compellent Threats US-imposed Economic Sanctions

10 MODEL SPECIFICATIONS/ ROBUSTNESS CHECKS Model: BTSCS and Probit; squared & cubed time trend terms, country fixed effects *States leave dataset after acquiring nuclear weapons, return if they give it up and becoming latent again Dependent Variables: Lagged by 5 years Covariates Military capability (CINC) Nuclear weapons acquisition Regime type (Polity, DCH5) Ally of the US Economic covariates (GDP per capita, economic openness) NPT Signatory, NPT Ratification Rivalry

11 FINDINGS Explanatory Variables Direction Significance 1) Military Assistance from U.S. Positive Not Significant 2) Economic aid from U.S. Negative Significant (***) 3) U.S.-imposed economic sanctions Positive Significant (***) 4) MIDs Initiator Positive Significant (***) 5) MIDs Target Positive Significant (*) 6) Compellent threat success Positive Not Significant With controls, FE *** p<0.01, ** p<0.05, * p<0.1

12 IMPLICATIONS Qualified support for latent provocation theory Some positive side-effects: Associated with military aid, success in issuance in compellent threats Some negative consequences: Lower incidence of economic aid, increased associated with economic sanctions Uncertain impact: Outcome when initiating disputes Latency may make states more risk-acceptant and revisionist.

13 CONCLUSIONS Iranian acquisition of latency may help it deter threats and obtain resources; may also reduce likelihood Iran aims for nuclear weapons if deterrent benefits are seen as sufficient Reactive spread of latent technology to other states that want to reap the benefits

14 BACK-UP SLIDES

15 TABLE 1: EFFECTS OF LATENCY ON DISPUTE AND BARGAINING OUTCOMES (BTSCS) Latency Model 1 (Mil Aid) Model 2 (Econ *** Aid) Model 3 (Sanctions) 1.338*** Model 4 (ThreatSuccess) Model 5 (MIDsTarg) 0.210* Model 6 (MIDsInit) 0.861*** (0.188) (0.129) (0.149) (0.532) (0.124) (0.122) NW Status ** (0.824) (0.584) (0.676) ( ) (0.502) (0.502) Economic Openness *** *** *** *** (0.002) (0.002) (0.004) (0.015) (0.002) (0.002) Regime Type *** *** *** (0.007) (0.006) (0.009) (0.031) (0.006) (0.006) NPT Signatory 0.275* 0.293*** *** ** (0.143) (0.097) (0.128) (0.591) (0.104) (0.108) GDP Per Capita 0.000*** *** *** *** (0.002) Polity Change ** * 0.081*** (0.010) (0.010) (0.013) (0.050) (0.010) (0.010) Rivalry 0.397*** *** 0.701* 1.129*** 1.562*** (0.086) (0.070) (0.104) (0.417) (0.078) (0.085) CINC *** *** 6.823* *** 8.202*** (3.664) (1.886) (1.479) (3.816) (1.656) (1.501) US Ally 1.267*** 0.892*** 0.530*** *** (0.100) (0.090) (0.127) (0.451) (0.089) (0.093) Time *** *** 0.230*** ** * (0.270) (0.017) (0.024) (0.110) (0.018) (0.019) Time *** 0.008*** *** 0.015** 0.002* (0.174) (0.001) (0.001) (0.006) (0.001) (0.001) Time *** *** 0.000*** *** (0.027) Observations 6,809 6,809 6,809 3,119 6,814 6,814 Standard Errors in Parentheses; *** p<0.01, ** p<0.05, * p<0.1

16 TABLE 2: CONDITIONAL EFFECTS OF NUCLEAR LATENCY ON ALLIES Latency Model 1 (Mil Aid) Model 2 (Econ * Aid) Model 3 (Sanctions) 0.775** (0.302) (0.243) (0.372) NW Status (1.452) ( ) ( ) Economic Openness ** *** *** (0.003) (0.003) (0.004) NPT Era 1.217* 1.874*** 3.882*** (0.466) (0.481) (0.500) Regime Type * *** (0.014) (0.016) (0.018) NPT Signatory *** (0.317) (0.261) (0.306) GDP Per Capita *** *** Polity Change ** ** 0.073*** (0.019) (0.021) (0.020) Rivalry 0.408*** 0.449*** 0.712*** (0.158) (0.164) (0.210) CINC *** ** ** (10.206) (8.504) (19.559) Time *** *** 0.425*** (0.585) (0.066) (0.135) Time *** 0.029*** (0.363) (0.006) (0.021) Time *** *** (0.055) (0.001) Observations 2,270 2,270 2,270

17 TABLE 3: CONDITIONAL EFFECTS OF NUCLEAR LATENCY ON NON-ALLIES Latency Model 1 (Mil 0.870*** Aid) Model 2 (Econ *** Aid) Model 3 (Sanctions) 1.989*** (0.291) (0.181) (0.180) NW Status (1.073) (0.715) (0.759) Economic Openness ** *** (0.001) (0.001) (0.002) NPT Era 0.827*** 1.302*** (0.234) (0.206) (0.217) Regime Type *** * (0.009) (0.006) (0.010) NPT Signatory 0.384** 0.422*** (0.158) (0.107) (0.155) GDP Per Capita *** *** Polity Change ** 0.066*** (0.013) (0.012) (0.017) Rivalry 0.213*** *** (0.109) (0.082) (0.127) CINC 7.102** (3.577) (1.972) (1.577) Time *** *** 0.150*** (0.297) (0.020) (0.028) Time *** 0.007*** *** (0.191) (0.001) (0.002) Time *** *** 0.000*** (0.030) (0.001) Observations 4,539 4,539 4,539