The Cost of Protection: Explaining the Success or Failure of the Burden-Sharing Pressure of the U.S. Alliance
Brian Blankenship is an assistant professor of political science at the University of Miami with research interests in international relations, international cooperation, and U.S. diplomacy.
来源：Brian Blankenship (2021) The Price of Protection: Explaining Success and Failure of US Alliance Burden-Sharing Pressure, Security Studies, 30:5, 691-724.
Guide
Why does the United States exert pressure on its allies to share responsibilities sometimes succeed and sometimes fail? How to explain this change? Most of the existing studies have focused on the characteristics of asymmetric alliances and the results of burden-sharing, and less attention has been paid to the issue of pressure-sharing by major powers on their allies. This paper argues that the ability of a patron state to pressure its allies to increase burden-sharing efforts depends on the extent to which the allies fear their withdrawal from the alliance. The authors argue that the strategic value of allies and the threat environment affect the credibility of patrons withdrawing from the alliance. When an ally has a low strategic value, it is easier for a patron to abandon that ally. The higher the strategic value of an ally, the more likely it is that the patron state will protect the ally. In terms of threat environment, the authors divide threat perception into four aspects: adversary capabilities, adversary behavior, adversary intentions, and geographic vulnerability. The author argues that if allies share common adversaries with patronic states, they are more likely to share responsibility. In terms of geography, the difference in the sharing of responsibilities depends on the distance of allies from these adversaries. In particular, the authors point out that the effects of proximity are not linear, and that countries that are adjacent on land with common adversaries are particularly vulnerable. Based on a series of theoretical derivations, the author proposes three (four) hypotheses. Hypothesis 1a (H1a): When allies and patrons have a common adversary, it is easier for allies to share responsibility. Hypothesis 1b (H1b): When allies share land borders with their adversaries, it is easier for allies to share responsibility. Hypothesis 2 (H2): Allies are unlikely to share responsibility if they are located near the sea choke point around their asylum adversary. Hypothesis 3 (H3): When an ally has fewer U.S. allies among its neighbors, it is less likely that the allies will share the blame. The authors' quantitative analysis then strongly demonstrated H1b, H2, and H3, while H1a received weak support. In terms of specific causal mechanisms, the authors not only use quantitative methods to distinguish the theory from alternative explanations by testing the surrogate dependent variable of host support of the allies, but also compare the cases of Japan and West Germany to demonstrate the impact of the fear of abandonment on the sharing of responsibility. Finally, based on the research in this paper, future research could further expand this topic by examining when asylum states ask for responsibility sharing, other means of responsibility sharing, and other reasons for the success or failure of responsibility sharing.
introduction
What would explain the success or failure of the U.S. shift in burden-sharing? This paper argues that the large differences in the burden of responsibility among allies can be explained by the ability of patrons to exploit the psychology of allies for fear of abandonment. The more credible the threat of asylum's abandonment, the more effectively the patron can encourage burden-sharing if the situation is worse for allies to be abandoned. The threat can be explicit or overt, implicit or private. This paper identifies structural conditions in which allies are more vulnerable to coercion on asylum and more likely to engage in responsibility-sharing. This paper argues that when allies have limited strategic value and asylum states have more abundant substitutions for local partners, asylum states are most effective in threatening to abandon allies. To test this theory, this paper uses a statistical analysis of the military support of U.S. allies from 1950 to 2010. There are two main contributions to this article. First, this paper fills in the gaps in the existing literature and pays attention to the bargaining process that shapes the sharing of responsibility in alliances. Second, the theory of this paper has implications for understanding how great powers manage their partnerships, deter adversaries, and maintain influence in the international system. This paper is organized as follows, first reviewing the existing literature on alliance responsibility sharing and proposing a theory. The paper then discusses the study design and policy implications and gives econometric results. Finally, a discussion of theoretical and policy implications and future research directions are given.
Burden-sharing for asymmetric alliances
Research on the sharing of responsibilities in asymmetric alliances can be broadly divided into two categories. The first group considers asymmetric alliances to be characterized by asymmetric commodity exchanges, i.e., large countries provide security and small countries provide autonomy. The second focuses on the outcome of the sharing of responsibilities and draws on the logic of collective goods. There is another alternative explanation that has not been adequately studied in existing studies, and that is great power pressure. Sanctuaries don't need to simply accept the level of effort of their allies and can try to convince them to do more. Many scholars have long suggested that powerful countries can force their allies to do more in the sharing of responsibilities. However, the conditions under which this pressure is effective have not yet been theorized and have not been the subject of empirical testing.
The theory of effective responsibility sharing of pressure
Shared responsibility is a core part of alliance management. While allies have an incentive to free ride, patronic states can use allies' fear of abandonment to temper that motivation. The leverage of the asylum can shape the responsibility sharing of the alliance in three ways. First, patronic states can actively exploit their allies' fear of abandonment through direct threats, either overtly or privately. Second, the asylum state can indirectly exploit the allied country's fear of abandonment, and while the patron state encourages the allied state to share the responsibilities of the ally, the allied country's fear of abandonment also plays a subtle role. Third, allies can take the initiative to share responsibilities, preemptively strike, hedge and defend against the threat of abandonment. One of the main strategies of the state to avoid being abandoned is to make itself more valuable. All three causal pathways can occur, so the authors ultimately don't know which causal pathways will work in a particular situation. But all three pathways point to the same empirical prediction that when the threat of U.S. abandoning allies is more credible, allies take on more shared responsibility. This article explores the conditions under which asylum states can more credibly make the sharing of responsibilities of their allies a condition for their protection.
The ability of a patron state to pressure its allies to increase responsibility-sharing efforts depends on the extent to which allies fear that they will withdraw from the alliance. As an alternative to their alliance, countries of asylum have two options. First, it can unilaterally renounce its commitments or reduce the protections it provides, whether pursuing a more self-sufficient, or even isolationist foreign policy, or seeking reconciliation with its adversaries. The two approaches are not mutually exclusive, but in fact mutually reinforcing; Second, the asylum country can find other allies. Thus, the extent to which a state of asylum is capable of taking either of these options determines the credibility of its threat to abandon its allies. Specifically, the authors argue that two factors, the strategic value of allies and the threat environment, influence the credibility of these external options.
When an ally has a low strategic value, it is easier for a patron to abandon that ally. The higher the strategic value of an ally, the more willing the patron will be to ensure that it remains in its camp, and therefore the more likely it is that the patron will protect the ally. At the same time, if the situation of the allies is worse without the protection of the asylum state, the fear of abandonment may also be more prominent. In particular, when allies have a higher degree of perception of external threats, allies are more worried about abandonment and thus more susceptible to pressure from patrons. The more vulnerable an ally is to a common adversary, the higher the cost of fighting without the assistance of the patronic state. Thus, the ability of a patron state to exploit an ally's fear of abandonment depends both on whether the ally considers its withdrawal threat credible and on the extent to which the ally relies on its protection. But these two factors are often intertwined, and many of the factors that make allies more vulnerable to outcasts can also reduce the credibility of the threat of a patron. The external threat posed by a common adversary not only makes allies dependent on the protection of patrons, but also makes allies invaluable in containing and deterring adversary expansion. Therefore, the challenge in making empirical predictions is that the common threat environment of the alliance is closely related to the strategic value of the allies to the asylum countries. To distinguish between these two mechanisms and predict their net impact on allied responsibility sharing, it is necessary to distinguish between the various components of threat perception: adversary capabilities, adversary behavior, adversary intentions, and geographic vulnerability. The authors argue that adversary capabilities and behaviors can shape threat perceptions of both allies and patrons, but that perceptions of geography and adversary intentions vary across alliances. The differences in perceptions of adversary intentions between patrons and allies are difficult to measure, and the factors that influence threat assessment are diverse, so this paper simply predicts that if allies and asylum countries share adversaries and perceive the adversary as a high level of threat, on average, those allies are more likely to share responsibility than those who do not have a common adversary with the asylum state and perceive the adversary as less threatening. This, in turn, demonstrates the importance of geography in the sharing of responsibilities in alliances, as geography is both easy to measure and varies from alliance to alliance. The proximity of a common adversary is both a source of vulnerability and a source of strategic value. For patrons, allies adjacent to their adversaries are a useful shield against their adversary's expansion before they become a greater threat in the future, as these allies can both physically block the adversary's expansion and provide the patron with territory from which they can project power to achieve the same end. However, for allies, the proximity to an adversary would expose them to a direct attack. Thus, the effectiveness of an ally's proximity to a common adversary depends on whether the ally is so vulnerable to attack that its vulnerability outweighs the value it brings to the patron state.
Based on the above discussion, the authors believe that two predictions can be made. First, allies who do not have a common adversary with the country of asylum are unlikely to share responsibility. Second, in countries with which they have common adversaries with asylum, the difference in the allocation of responsibilities depends on their proximity to those adversaries. The authors further point out that the effects of proximity are not linear, and that countries that are adjacent on land to a common adversary are particularly vulnerable, and that, despite their strategic value, they are likely to be affected by the burden-sharing pressures of their asylum states. Countries that share maritime borders with common adversaries do not spend more or less on defense than their more distant allies. So hypothesis 1 is as follows,
Hypothesis 1a (H1a): When allies and patrons have a common adversary, it is easier for allies to share responsibility.
Hypothesis 1b (H1b): When allies share land borders with their adversaries, it is easier for allies to share responsibility.
In addition, the author believes that when allies are close to key sea chokepoints around their opponents, they are unlikely to share responsibility. Because these allies play a key role in cutting off the power projection of their opponents.
Hypothesis 2 (H2): Allies are unlikely to share responsibility if they are located near the sea choke point around their asylum adversary.
The last element of the strategic value of an ally is the abundance of alternative options for patronic states. All other things remain the same, and the patron state is less able to abandon an isolated ally than an ally surrounded by its neighbors, who are also allies of the United States. As a result, the author expects less defense spending from nearby allies of the United States, which have fewer allies.
Hypothesis 3(H3): When an ally has fewer U.S. allies among its neighbors, it is less likely that the allies will share the blame.
Alternative explanations
The theoretical predictions in this paper are contrary to the following alternative explanations. The first is the "economic theory of alliances," which predicts that larger allies will contribute more GDP to defense. The second is the type of regime, which believes that democracies are more accountable to their people than authoritarian states, and some studies have shown that democracies spend less on national defense. Third, current research suggests that the presence of U.S. troops should be associated with lower defense spending by allies. However, the theory of this paper is that the level of U.S. military presence is often an endogenous consequence of other factors such as a shared threat environment or the strategic value of allies. In addition, U.S. policymakers can use the threat of reducing troop numbers to encourage shared responsibility. Finally, traditional balance-of-power theories predict that an ally's defense efforts are an internal counterweight to their threat environment. The theory of this paper does not deny this view, but it argues that external threats affect another mechanism of allied responsibility-sharing: they make allies fear abandonment and make their coercion of patrons more vulnerable. And the theory of this paper makes a novel empirical prediction: an external threat will affect the allocation of responsibility only if the strategic value provided by the allies is insufficient to offset the external threat, that is, when the allies have a land border with a common adversary. Indicators such as adversary capacity and behaviour may have no net effect on the allocation of responsibility, as they weaken the threat of abolition by making allies more vulnerable to abandonment and making allies more valuable to asylum recipients. In contrast, the balance of power theory predicts that all threat indicators will have a consistently positive impact on allied military spending.
Study design
This paper uses a cross-border dataset of U.S. allies to share responsibilities to test hypotheses. The unit of analysis is the US allies - years. The sample for this article covers the period from 1950 to 2010 and includes all countries that have defense agreements with the United States, as defined by the Database of War-Related Factors (COW). Then, the nature of the Rio Treaty (not an agreement to counter external threats but to manage conflicts between member states), the nature of the U.S. relationship with countries in the region (the U.S. sees the Americas as its sphere of influence, and the threat of abandoning allies is not prominent in these alliances), and the lack of variation in the independent variables of this paper, this paper excludes U.S. allies in the Americas.
Dependent vs. independent variables
The primary dependent variable for this paper is allied military spending as a percentage of GDP, and the data is derived from version 5.0 of the National Material Capabilities (NMC) dataset of the War-Related Factors database. It is difficult to measure the threat of abandonment directly because they are rarely overt and often indirect. And the allies may be reluctant to openly clash so as not to give the adversary the impression that the alliance is not united, or to cause domestic dissatisfaction with the alliance. Still, if the assumptions of this paper are correct, we will still see systemic differences in allied military spending, based on how allies score on independent variables, whether due to direct or indirect patron pressure, or due to allies actively increasing defense spending for fear of being abandoned.
The first set of independent variables in this paper uses two dummy variables to capture the presence and adjacency of a common adversary: if the ally has a land border with a common adversary, the dummy variable has a value of 1, otherwise it is 0, and if the ally has a common adversary with the United States but no land border, the other dummy variable has a value of 1, otherwise it is 0. This paper expects both to have a positive impact on allied military spending, and that the former will have a greater impact, as allies bordering a common adversary on land are more vulnerable and therefore particularly vulnerable to U.S. burden-sharing pressures. In some models, dummy variables are also added to the question of whether allies are 400 miles or less away from a common adversary, and whether an ally has a common adversary with the U.S. but is not close to the former. The expectation of this paper is that the impact of both on the sharing of responsibilities among allies is about the same.
Common adversaries have the following definitions. First, any country that is hostile in foreign policy with the United States and its allies, including Russia (1950-89, 2007-2010), China (1950-72, 1996-2010), and Cuba (1959-2010), but this excludes many other countries that can be considered common adversaries, so this article also codes any countries in the Soviet bloc during the Cold War as common adversaries. Since the post-Cold War era lacks an overarching threat to America's allies, the threat posed by allies to the United States has become more varied, so this article also includes Iraq, Iran, and North Korea among America's adversaries. Third, we set up a dummy variable of 1 if the ally is within 400 miles of a major maritime choke point near a major U.S. adversary, and ships need to pass through the choke point to reach open water. In practice, the article focuses on the choke points that encircled the Soviet Navy during the Cold War, and the choke points that encircled the Russian and Chinese navies after the Cold War, because during the study their power projection capabilities were unique among American competitors. This does not mean that this article denies the strategic value of other choke points, but rather that the value of these "front-line" choke points lies in the fact that they can prevent the adversary from projecting power further away from the start. Finally, this article measures the number of other allies in the vicinity by the number of allies adjacent to each one, whether by land or by water of four hundred miles or less.
Control variables
This paper first controls the annual GDP growth rate of allies (measured as a percentage of the previous year's GDP) because we expect that growing countries will collect more taxes for defense spending, while countries with declining economies will have less government revenues. Second, this paper controls the size of the allies' economies, using the Allies' GDP in 2005 constant dollars to control the size of the allies' economies, which the available literature anticipates will have a positive impact on the allies' defense responsibilities. Third, this paper uses the binary indicator of democracy to control the type of regime of allies. In addition, in terms of the security environment, the authors consider the unique conflict environment of each ally, controlling the number of militarized interstate disputes (MIDs) in which the allies are involved, weighted according to the degree of hostility of the MIDs, and the number of allies in foreign policy hostile to the United States. Secondly, the article also includes the dummy variable of whether the allies have a colonial empire. Thirdly, the author exercised control over the number of American troops in the territory of each allied country. Finally, the author also included in some models the minimum distance of the allies from the United States.
Model description
In this paper, the model is estimated using ordinary least squares regression, and the standard errors are grouped by country. In addition, regional fixed effects and, in some cases, year fixed effects are added to account for unobserved regional and temporal variations.
Results & Robustness
The results in Figure 1 provide strong support for the hypothesis of this paper. In general, H1b (land proximity to common adversaries), H2 (proximity to sea chokepoints) and H3 (alternative U.S. allies) are strongly confirmed. The results of the study consistently show that those that have a common adversary but no land adversary, or no common adversary, spend a higher percentage of their GDP on defense than those that have a land adversary. In addition, allies close to strategically valuable sea chokepoints, as well as allies with fewer other U.S. allies in the vicinity, spend less on defense. H1a (a common adversary with no land adjacency) has weak support and only has a significant positive effect at the 0.1 level in some models. In this paper, two main variables, the proximity to the sea choke point and the common adversary of the United States, are broadly and narrowly coded to test the robustness of the research results, and the results are consistent with the original results.
Figure 1: Key results
Note: Standard errors by country are in parentheses. FE: Fixed effect.
p < 0.10, *p < 0.05, ** p < 0.01, *** p < 0.001
The dependent variable is the military spending of the allies (as a percentage of GDP)
Exploring causal mechanisms: the fear of abandonment vs. internal checks and balances
The potential challenge in evaluating this theory is that the predictions for H1a and H1b are consistent with those based purely on allies' perceptions of external threats. To distinguish the theory from alternative explanations, this paper first uses an alternative dependent variable, allies' host-nation support (HNS) for the U.S. military presence in their countries. Because this is an ally's host support as a compensation for the United States, it is more likely to be the result of U.S. pressure than a response to internal checks and balances, as shown in Figures 2 and 3, which turn out to be more host support from countries that share the U.S. adversaries and are adjacent to its land. Second, this paper provides qualitative evidence from the 1960s to the 1970s in West Germany and Japan. In terms of case selection, this paper tries to focus on changes in the threat environment and try to control for other changes, the two countries are similar in size, type of regime, strategic value, World War II legacy that makes the United States fear a resurgence of militarism, and both have more troops than other U.S. allies. The period was chosen because the economies of America's allies were recovering, but at the same time America's dominance was weakening, the Vietnam War was limiting America's resources, and U.S. officials were under pressure from Congress to share more responsibility. By comparing the cases of Japan and West Germany, the authors find that West Germany's fear of abandonment was higher than Japan's, and that this fear led to its defense efforts in response to more American pressures to share responsibility.
Figure 2: Results of Allied Support to the Host Country.
The dependent variable is each ally's support for the host country (as a percentage of GDP)
Figure 3: Results of Allied support for the host country.
The dependent variable is the residuals of each ally's support for the host country.
Implications and paths for future research
This study provides many avenues for future research. First, although this study is primarily concerned with the impact of fear of abandonment as a source of disparity in responsibility sharing, further research could investigate the effectiveness of other means. These include, but are not limited to, economic coercion, inducements, and "naming and shaming" of under-burdened countries. A related line of research could focus on the willingness of allies to contribute to the collective good for reasons unrelated to bargaining or other allied contributions, such as normative considerations and values. In addition, because the independent variables in this paper are relatively fixed from year to year, more research could be done on when a state of asylum really wants its allies to take more responsibility for their own defense, and how to interpret the changes that have occurred over time.
Finally, this paper has the following implications for the discussion of the grand strategy of the United States. This article argues that if U.S. officials want their allies to take more responsibility, they should be interested in their allies' fear of abandonment. The author argues that even if the United States goes to great lengths to reassure its allies, it is not helpless in influencing allies to share the burden, as it can effectively combine pacification with the fear of abandonment. Thus, the trade-off between appeasing allies and encouraging burden-sharing is not absolute. However, the US maneuver will be limited by factors beyond its control. And the United States can also weaken the sharing of responsibilities by expanding the alliance front and isolating larger allies from common adversaries (e.g., NATO's eastward expansion after the Cold War). Moreover, a shared external threat does not guarantee success, and while increased awareness of the threat by allies may favor greater sharing of responsibility, increased awareness of the threat by the United States may discourage allies from sharing more responsibility. The trends of the last decade have been the relative weakening of American power and the rise of China. If the U.S. has fewer resources available for foreign commitments, and allies are concerned about China's rise, they may be willing to share some of the blame. However, the extent of U.S. success may be influenced by its own perception of the threat, thus reducing the incentive for allies to share responsibility.
Vocabulary accumulation
Rio Treaty
Rio Pact
economic theory of alliances
The Economic Theory of Alliance
National Material Capabilities, NMC
The country's material capacity
year fixed effects
Year-to-year fixed effect
allies’ host-nation support, HNS
Allies of the host support
naming and shaming
Point Name Insult
Translator: Li Kailang is a national politics scholar and compiler, Waseda University's Graduate School of Asian and Pacific Studies, with research interests in alliance politics, East Asian security, and international relations theory.
Proofreading | Li Linjie Deng Haoran
Audit | Ding Weihang
Typography | Han Xinjie
This article is a public welfare sharing, serving scientific research and teaching, and does not represent the views of this platform. If there is any omission, please correct it.