Sage Journals: Discover world-class research

Abstract

Leeds et al. (2000) report that military alliance commitments are honored in war around 75% of the time. We update and extend data on alliance reliability from 1816 to 2003. Our analysis reveals a lower compliance rate overall: 50%. We find a sharp disparity in alliance reliability before and after World War II. States honored their alliance commitments 66% of the time prior to 1945 but the compliance rate drops to 22% from 1945 to 2003. Moreover, the rates of fulfillment for defense pacts (41%) and nonaggression pacts (37%) are dramatically lower than offensive alliances (74%) and neutrality agreements (78%). These findings carry implications for the role of military alliances in world politics and highlight the need for more research to explain the differences that emerge before and after World War II.

Keywords

Military alliances treaty fulfillment war

An influential study by Leeds, Long, and Mitchell (2000, hereafter “LLM”) reports that countries honor their military alliance commitments 74.5% of the time. The view that states uphold their alliance promises in war the vast majority of the time has now become conventional wisdom in scholarship (e.g. Guzman, 2008: 11; Leeds, 2003a: 803; Simmons, 2010: 280–81). LLM focus on the period from 1816 to 1944, raising the possibility that a different pattern might exist in the post-World War II era. We follow their procedures to produce a dataset on alliance treaty reliability over a longer period – from 1816 to 2003 – drawing on the Alliance Treaty Obligations and Provisions (ATOP) dataset (Leeds et al., 2002) and updated information on participation in interstate wars (Reiter et al., 2014, Version 1.1).

Our analysis reaffirms LLM’s finding that alliance treaties were fulfilled the majority of the time from 1816 to 1944. After World War II, however, countries honored their alliance promises in just 22% of cases in which treaty commitments were invoked in war. Overall, we find that alliances were upheld 50% of the time in interstate wars from 1816 to 2003. Our updated analysis therefore shows that alliance reliability rates are lower than previously believed, and that there is a sharp disparity in promise fulfillment in the pre- and post-World War II periods. These findings carry implications for the rich literature on the sources and effects of alliance treaty fulfillment in international relations (e.g., Crescenzi et al., 2012; Gartzke and Gleditsch, 2004; Gibler, 2008; Leeds et al., 2009; Leeds and Savun, 2007; LeVeck and Narang, 2017).

Alliance reliability in war: Data and procedures

Scholars have discussed the honoring of alliance promises at least since the time of Thucydides (431 BCE). Systematic research on this topic began to take off after scholars compiled comprehensive data on military alliances and war through the Correlates of War project. Early studies were not sanguine about the reliability of alliances in war: Sabrosky (1980) finds that alliances are honored less than 30% of the time, while Siverson and King (1980: 2) conclude that just 23.1% of the opportunities to fulfill an alliance commitment resulted in countries joining their allies in war. One limitation of these studies, however, is that they do not consider the specific requirements of each alliance (i.e., the casus foederis). This research therefore misclassifies alliance fulfillment: some treaties are considered to be violated (or upheld) even though countries did not renege on (or honor) any particular promise.

Leeds et al. (2002) address this issue by developing the ATOP dataset, which identifies specific alliance provisions to more accurately assess when states are (and are not) obligated to act. LLM demonstrate that accounting for the specific requirements of military alliances leads to a much more optimistic view of alliance reliability in war. While their study generates a more accurate assessment of wartime alliance reliability in the pre-1945 period, it leaves a key question unanswered: what is the rate of treaty compliance after World War II, and how similar is this rate to the pre-war period?

To address this issue, we create a dataset of alliance treaty fulfillment for the period from 1816 to 2003, following the procedures described in LLM. In addition to examining a longer time period, we depart from LLM in one key respect: we use updated war data compiled by Reiter et al. (2014). Otherwise, we mimic LLM’s research design.

The unit of observation in our dataset is the “alliance performance opportunity.” Because some wars trigger multiple alliance commitments, the same conflict may appear in the dataset more than once. The dataset accounts for four types of alliances: defense pacts, offensive alliances, neutrality agreements, and nonaggression pacts. We started by identifying all active alliance commitments made by participants in 70 interstate wars during our period of study. Next, we determined whether each treaty’s casus foederis was met in a given war. The analysis ultimately includes only those cases where the particular requirements of the treaty were met, excluding cases that are not relevant for assessing alliance reliability. We identified a total of 576 “alliance performance opportunities” from 1816 to 2003. The casus foederis was invoked in 146 of these cases, compared to 110 with LLM.¹

For all “alliance performance opportunities” in which the casus foederis was invoked, we determined whether the commitment was upheld. Like LLM (693–94), we classify a commitment as upheld if: “(1) an ally fought on the same side with its alliance partner as promised (in a defense and/or offense pact), (2) an ally remained neutral in a conflict as promised, or (3) an ally fought alongside its alliance partner despite promising only neutrality.” A treaty promise is violated if “(1) an ally fought against its partner (in the context of a defense, offense, nonaggression, or neutrality pact), or (2) an ally did not come to a partner’s aid even though it had promised such assistance in a defense and/or offense pact” (LLM: 694). Some alliances are multilateral. For these cases, researchers must decide how to code alliance reliability when some allies fulfill a commitment but others do not. Sabrosky (1980) classifies an alliance as fulfilled if any ally honors a promise, while LLM code a promise as upheld only if all partners honor their treaty-backed pledges. We adopt the same standard as LLM.²

Findings

We found that alliance treaties were fulfilled in war from 1816 to 2003 at a rate of 50% (see Figure 1).³ This figure is about 25 percentage points lower than the alliance reliability rate reported by LLM. Thus, updating the dataset through the post-World War II period generates a sizable change in the overall compliance rate.

Figure 1.

Updated data on alliance treaty fulfillment.

There is interesting variation in the rate of compliance based on treaty type.⁴ Offensive alliances and neutrality agreements are fulfilled at high rates − 73.81 and 77.78%, respectively. Defense pacts, by contrast, are upheld just 41.18% of the time. Nonaggression pacts are honored at the lowest rate (36.73%). The type of alliance commitment is therefore important for understanding promise fulfillment. This may be partially because defense pacts and nonaggression pacts are more likely to generate time inconsistency problems. These alliances may be more likely to be invoked years after they are signed, increasing the likelihood that the circumstances that led to forming the agreement have changed.⁵ On the other hand, offensive treaties and neutrality pacts seem to be forged shortly before a war begins, after allies have already decided to fight together (or to remain neutral).

The disparity in alliance treaty fulfillment before and after World War II is also striking (see Figure 2). From 1816 to 1944, we find that alliance treaties were upheld 66.30% of the time. This figure is similar to the rate of 74.5% in LLM.⁶ Our analysis confirms that alliances were reliable the majority of the time in the pre-1945 period. The story changes after World War II: from 1945 to 2003, alliances were upheld at a rate of 22.22%. States were therefore about three times as likely to honor alliance commitments from 1816 to 1944 as they were from 1945 to 2003.

Figure 2.

Comparing alliance treaty fulfillment in two eras.

Table 1 lists the opportunities to fulfill alliance commitments from 1945 to 2003. This information offers some possible explanations for the sharp decrease in promise fulfillment after World War II. One thing that stands out is that offense pacts and neutrality agreements, which are honored at higher rates than defense pacts and nonaggression pacts, are less common in the modern era. Only 20% of alliance performance opportunities from 1945 to 2003 included offensive or neutrality-based commitments, compared with more than 70% from 1816 to 1944. Changes in the types of alliance commitments that are invoked in war over time might partially account for the puzzling trend that we identify. Yet we still see sizable variation in compliance within the same commitment type in the two eras. Defense pacts, for instance, were honored 61.02% of the time from 1816 to 1944 and only 13.95% of the time post-World War II. What else, then, could explain the changing patterns over time?

Table 1.

Alliance performance opportunities, 1945–2003.

War	ATOP ID	Allies	Commitment Type	Honored or Violated
World War II	3030	China, USSR	D,O	Honored
World War II	3010	Multilateral (Americas)	D	Violated
World War II	3020	USSR, Yugoslavia	D,O	Violated
World War II	3005	USSR, UK, USA	O	Honored
Korean	3215	Australia, New Zealand, USA	D,NA	Violated
Korean	3210	Philippines, USA	D,NA	Honored
Korean	2170	Afghanistan, USSR	N,NA	Honored
Taiwan Strait I	3270	Taiwan, USA	D,NA	Violated
Sinai	3305	Egypt, Saudi Arabia	D	Violated
Sinai	3205	Multilateral (Arab)	D,NA	Violated
Sinai	3322	France, Israel, UK	O,N	Honored
Sinai	3310	Egypt, Saudi Arabia, Yemen	D	Violated
Sinai	3300	Egypt, Jordan, Syria	D	Violated
Soviet–Hungarian	2170	Afghanistan, USSR	N,NA	Honored
Soviet–Hungarian	3285	Multilateral (Warsaw)	D,NA	Violated
Soviet–Hungarian	3110	Hungary, Romania	D,NA	Violated
Soviet–Hungarian	3120	Hungary, USSR	D,O,NA	Violated
Taiwan Strait II	3270	Taiwan, USA	D,NA	Violated
Bizerte	3395	CAR, Chad, Congo, France	D	Violated
Bizerte	3390	France, Madagascar	D	Violated
Bizerte	3410	France, Gabon	D	Violated
Malaysia	3330	Malaysia, UK	D	Honored
Six Day	3205	Multilateral (Arab)	D,NA	Violated
Six Day	3570	Egypt, Iraq, Jordan	D	Violated
Six Day	3467	Egypt, Yemen	D	Violated
Six Day	3538	Egypt, Syria	D	Honored
Six Day	3465	Jordan, Saudi Arabia	D,O	Violated
Six Day	3350	Egypt, Iraq	D	Violated
Football	3075	Multilateral (Rio Pact)	D,NA	Violated
Football	3150	Multilateral (OAS)	D,NA	Violated
Football	3490	El Salvador, Guatemala, Honduras, Nicaragua	D	Violated
Bangladesh	3675	India, USSR	N,NA	Honored
Bangladesh	3260	Multilateral (SEATO)	D,NA	Violated
Bangladesh	3530	India, Pakistan	NA	Violated
Bangladesh	3355	Pakistan, USA	D	Violated
Yom Kippur	3465	Jordan, Saudi Arabia	D,O	Honored
Second Ogaden	3205	Multilateral (Arab)	D,NA	Violated
Second Ogaden	3485	Ethiopia, Kenya	D	Violated
Ugandan–Tanzanian	3700	Sudan, Uganda	D	Violated
Lebanon	3205	Multilateral (Arab)	D,NA	Violated
Lebanon	3807	Algeria, Libya, Syria, Yemen	D	Violated
Aouzou Strip	3205	Multilateral (Arab)	D,NA	Violated
Persian Gulf	3015	Multilateral (Arab)	NA	Violated
Persian Gulf	4020	Canada, Bahrain	D	Honored
Persian Gulf	3205	Multilateral (Arab)	D,NA	Violated
Persian Gulf	3997	Iraq, Saudi Arabia	NA	Violated
Azeri–Armenian	4220	Multilateral (post-Soviet)	D,NA	Violated
Azeri–Armenian	4550	Azerbaijan, Turkey	D	Violated
Azeri–Armenian	4400	Multilateral (post-Soviet)	D,NA	Violated
Azeri–Armenian	3740	Multilateral (Helsinki Accords)	NA	Violated
Badme	3485	Ethiopia, Kenya	D	Violated
Kosovo	4820	Romania, Serbia	N,NA	Honored
Kargil	3705	India, Pakistan	NA	Violated
Iraq	3205	Multilateral (Arab)	D,NA	Violated

Notes: D=defense pact; O=offense pact; N=neutrality pact; NA=nonaggression pact.

Alliances with more than four members are listed as “multilateral.”

The invention of nuclear weapons offers one potential answer. In the post-World War II era, the world’s most powerful states possessed nuclear arsenals. Given the destructive power of these weapons, alliance commitments from nuclear powers – especially defense pacts – have rarely been challenged in war (Fuhrmann and Sechser, 2014). As a result, major powers appear just a few times in Table 1 and relatively weak states account for a sizable percentage of all “alliance performance opportunities” from 1945 to 2003. This is relevant because state power may affect alliance treaty fulfillment. Countries should be most likely to intervene in war when their participation might change the outcome (Morrow, 1994; Smith, 1995). Relatively weak states, according to this line of thinking, are less likely to fulfill their alliance promises because they lack the material capacity to sway a conflict’s outcome.⁷ Israel probably would have prevailed in the Six Day War, for instance, even if Saudi Arabia had joined its allies in the fight. From Riyadh’s perspective, this may have made it easier to largely remain on the sidelines. This is, however, merely a preliminary conjecture.

Conclusion

This study presented an updated assessment of the reliability of military alliances. Our analysis, which extended earlier work to include the post-World War II era and utilized updated war data, yields three main conclusions. First, the overall rate of alliance fulfillment in war is lower than previously reported. Second, defense pacts and nonaggression pacts are honored much less frequently than neutrality agreements and offensive alliances. Third, there is a large disparity in the rate of alliance fulfillment before and after World War II. This offers a lesson for researchers that extends beyond the realm of alliance politics: trends that apply in one period may not extend to other eras. It is important to consider whether relationships of interest vary over time – particularly when there are structural shocks, like World War II.

What implications does the lower compliance rate carry for our understanding of alliance treaty reliability? Our analysis does not imply that military alliances are ineffective, nor does it challenge the evidence that defense pacts promote peace through extended deterrence (Leeds, 2003b; Johnson and Leeds, 2011; Fuhrmann and Sechser, 2014). The most effective threat is one that never has to be implemented (Schelling, 1966). NATO does not appear in our dataset, for example, precisely because potential adversaries perceive it as effective. However, when alliance commitments are invoked in war, allies uphold their promises less often than the conventional wisdom suggests. This implies that leaders are less restrained by treaty commitments than prior research would expect, a conclusion that carries implications for our understanding of international institutions (for a review of relevant literature, see Simmons, 2010).

Our analysis opens up avenues for future research. Scholars could use our updated dataset to revisit enduring debates about alliance politics, such as whether democracies make more (or less) reliable allies (c.f., Gartzke and Gleditsch, 2004; Leeds, 2003a). The disparity in compliance rates over time also represents a puzzle worthy of examination in scholarship. Something may have changed after 1944 that fundamentally altered the nature of alliance politics. We have speculated about the sources of this variation, but further analysis is necessary in order to achieve more definitive answers. The disparity in compliance across commitment types is worthy of further investigation in scholarship as well. Dedicated studies on why offense pacts are so much more reliable than defense pacts would be especially welcome. More generally, future research might consider the implications of our findings for extended deterrence, war-fighting, and the efficacy of international institutions.

Supplemental Material

Appendices2018-0223_(1) – Supplemental material for Reassessing the fulfillment of alliance commitments in war

Supplemental material, Appendices2018-0223_(1) for Reassessing the fulfillment of alliance commitments in war by Molly Berkemeier and Matthew Fuhrmann in Research & Politics

Footnotes

Acknowledgements

The authors’ names are listed alphabetically. The authors are indebted to Ashley Leeds for detailed and constructive feedback on this project. The authors also thank Jeffrey Arnold, Rosella Cappella, Scott Cook, Bryan Early, Erik Gartzke, Kristian Gleditsch, Florian Hollenbach, Jeffrey Kaplow, Michael Koch, Quan Li, Sara Mitchell, Mitchell Radtke, Dan Reiter, Ahmer Tarar, participants in a research workshop at the University at Albany – SUNY, and audience members in conference presentations at the annual meetings of the American Peace Science Association, the Peace Science Society (International), and the International Studies Association for helpful comments on earlier versions of this paper.

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.

Supplementary material

The supplementary files are available at

The replication files are available at: .

Notes

Carnegie Corporation of New York Grant

This publication was made possible (in part) by a grant from the Carnegie Corporation of New York. The statements made and views expressed are solely the responsibility of the author.

References

Crescenzi

MJC

Kathman

Kleinberg

Wood

(2012) Reliability, reputation, and alliance formation. International Studies Quarterly 56(2): 259–274.

Fuhrmann

Sechser

(2014) Signaling alliance commitments: Hand-tying and sunk costs in extended nuclear deterrence. American Journal of Political Science 4(58): 919–935.

Gartzke

Gleditsch

(2004) Why democracies may actually be less reliable allies. American Journal of Political Science 48(4): 775–795.

Gibler

(2008) The costs of reneging: Reputation and alliance formation. Journal of Conflict Resolution 52(3): 426–454.

Guzman

(2008) How International Law Works: A Rational Choice Theory. Oxford, UK: Oxford University Press.

Johnson

Leeds

(2011) Defense pacts: A prescription for peace? Foreign Policy Analysis 7(1): 45–65.

Leeds

(2003a) Alliance reliability in times of war: Explaining state decisions to violate treaties. International Organization 57(4): 801–827.

Leeds

(2003b) Do alliances deter aggression? The influence of military alliances on the initiation of militarized interstate disputes. American Journal of Political Science 47(3): 427–439.

Leeds

Long

McLaughlin Mitchell

(2000) Reevaluating alliance reliability: Specific threats, specific promises. Journal of Conflict Resolution 44(5): 686–699.

10.

Leeds

Savun

(2007) Terminating alliances: Why do states abrogate agreements? Journal of Politics 69(4): 1118–1132.

11.

Leeds

Ritter

McLaughlin Mitchell

Long

(2002) Alliance treaty obligations and provisions, 1815–1944. International Interactions 28(3): 237–260.

12.

Leeds

Mattes

Vogel

(2009) Interests, institutions, and the reliability of international commitments. American Journal of Political Science 53(2): 461–476.

13.

LeVeck

Narang

(2017) How international reputation matters: Revisiting alliance violations in context. International Interactions 43(5): 797–821.

14.

Morrow

(1994) Alliances, Credibility, and Peacetime Costs. Journal of Conflict Resolution 38(2): 270–297.

15.

Reiter

Stam

Horowitz

(2014) A Revised Look at Interstate Wars, 1816–2007. Journal of Conflict Resolution 10(1): 61–80.

16.

Sabrosky

(1980) Interstate alliances: Their reliability and the expansion of war. In: Singer

(ed) The Correlates of War II, Testing Some Realpolitik Models. Glencoe: Free Press, pp. 161–198.

17.

Schelling

(1966) Arms and Influence. New Haven, CT: Yale University Press.

18.

Simmons

(2010) Treaty Compliance and Violation. Annual Review of Political Science 13(1): 273–296.

19.

Siverson

King

(1980) Attributes of national alliance membership and war participation, 1815–1965. American Journal of Political Science 24(1): 1–15.

20.

Smith

(1995) Alliance formation and war. International Studies Quarterly 39(4): 405–425.

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.

0.00 MB

0.15 MB