Arms, Alliances, and Patron-Client Relationships

To the Editors (Tongfi Kim writes)

In "To Arm or to Ally?" Keren Yarhi-Milo, Alexander Lanoszka, and Zack Cooper present an elegant and powerful theory that explains conditions under which a great power (in their case, the United States) offers a client state arms, an alliance commitment, or both.¹ Their article does not give due attention, however, to what motivates a patron to provide security assistance to clients in the first place—especially its desire to influence its clients to obtain concessions in military, political, economic, and other policies.² Below I explain how Yarhi-Milo, Lanoszka, and Cooper's conceptualization of key variables inadvertently plays down the role of influence-seeking in U.S. policy.

First, Yarhi-Milo, Lanoszka, and Cooper's conceptualization of arms transfers does not reflect the wide range of ways in which a patron can support a client and the significant differences among the types of arms transfers with regard to buying influence. Their conceptualization of arms transfers, in which "a state gives another state weapons to augment its military capabilities" (p. 95), "covers multiple methods of provision, including sales, grants, and loans" (p. 97). As they discuss at the beginning of the article, however, the United States spends large sums of money on security assistance worldwide, a significant portion of which is spent on things other than U.S. arms (p. 91). Israel, for example, has been allowed by the U.S. government to use about 26 percent of its U.S. Foreign Military Financing to purchase Israeli weapons.³ Pakistan, in addition to Foreign Military Financing, had received about $13 billion by 2015 in Coalition Support Fund assistance for "operational and logistical support of U.S.-led counterterrorism operations."⁴ Because military aid other than arms transfers can increase the military capabilities of clients, it deserves analysis. At the same time, it is problematic not to differentiate sales, grants, and loans in analyzing transactions between allies. For instance, Japan has received more arms from the United States than Israel, but letting Japan buy U.S. arms with Japanese money is much less of a patron's favor to a client than giving Israel money to buy U.S. arms.⁵

Second, Yarhi-Milo, Lanoszka, and Cooper define "commonality of security interests" as "the extent of the threat that the client's primary adversary poses to the patron's core security interests" (p. 98), but their definition neglects the importance of the patron and client agreeing on how to deal with the adversary. For instance, in the early years of the Cold War, South Korea's and Taiwan's adversaries were also U.S. adversaries, but these clients' preference for offensive strategies posed significant problems for Washington. The United States, therefore, needed to exert influence on both states. In fact, one of the major motivations for the U.S. government in offering an alliance commitment to South Korea was to gain influence over its president, Syngman Rhee, who had tried to sabotage the Korean War armistice.⁶ Even when a patron shares an adversary with a client, influencing the client remains an important objective.

Third, Yarhi-Milo, Lanoszka, and Cooper argue that the perceived military balance between the client and its main adversary affects the patron's willingness to provide costly arms. They find that policymakers who argued that the military balance did not favor a client or that it was shifting against it advocated costly transfers. It is also plausible, however, that those who would like to support a client use the military balance as a justification. Since the policy transition to costly arms transfers, Israel has received large amounts of military aid from the United States despite Israel's military superiority against its adversaries. It is unclear how much of an advantage would be enough for the United States to reduce this aid. Meanwhile, some shifts in which states receive more U.S. military aid—for example, Pakistan after the terrorist attacks of September 11, 2001—seem to have little to do with the...