Inherent View of National Interest

Domestic Politics as National Interest

Contrary to an inherent view of national interest, others have claimed that an issue’s value is malleable and heavily driven by domestic elites. For example, scholars have explored which specific policies are implemented when certain coalitions are in power (Brooks Reference Brooks2013; Narizny Reference Narizny2003a; Reference Narizny2007; Trubowitz Reference Trubowitz1998). These policies, however, are often domestically contentious or polarizing because they tend to reflect the interests of a specific group. This is contrary to the concept of national interest that this article is trying to explore, which are issues that can receive widespread consensus about their importance across coalitions.

Another group of scholars has examined how sectoral interests manage to co-opt the national interest and rationalize it to the public (Hobson Reference Hobson1902; Snyder Reference Snyder1993). These studies are closer to the article’s focus, but as Snyder himself acknowledges, justifying select policies to the public is harder when there are diverse interests in the society. This article expands on this existing literature to explore how some issues have an easier time remaining justified as national interests when faced with multiple discourses about what is important to the nation.

Finally, scholars have also argued that leaders may have incentives to ramp up the value of an issue in order to gain domestic support or gain leverage in international bargaining (Mueller Reference Mueller1973; Weiss Reference Weiss2013). However, while leader cues would certainly influence public opinion, a leader-driven approach cannot tell the full story. Leaders are often criticized for backing down in conflicts or for escalating unnecessarily (Kertzer and Zeitzoff Reference Kertzer and Zeitzoff2017; Tomz Reference Tomz2007), and their attempts to frame an issue as important or unimportant are not always successful (Druckman Reference Druckman2001; Krebs Reference Krebs2015; Zellman Reference Zellman2015). In fact, leaders do not enjoy a completely free hand in portraying issues as national interests even in many authoritarian regimes (Li and Chen Reference Li and Chen2021; Weeks Reference Weeks2012). Therefore, the question of which issues have an easier time being widely framed as the national interest and why still remains. In the next section, I investigate this question by turning to a domestic distributional approach, examining what kinds of issues are more likely to resonate with the public as having important consequences for the entire nation.

Benefit Diffuseness and Perception as National Interest

I start by hypothesizing that issues perceived as providing widespread benefits to nation members are more likely to be regarded as being in the national interest than those providing concentrated benefits to specific domestic groups. This can be for both reasons of self-interest and normative concerns. First, the public may favor issues providing widespread benefits because they believe they have a higher chance of directly gaining from the issues than from issues with concentrated benefits. By contrast, issues with concentrated benefits would only appeal to a small fraction of the society. Although those who gain directly from the concentrated benefits may still consider the issue as important and worth fighting for, if a vast majority of the public who do not directly gain do not think of the issue as important, there would be lower public support for the issue on average. Support for the issue would also be more nationally divisive, since some would perceive the issue to be important while others would not.

Second, people may have a normative understanding of how national interest should benefit the general citizenry rather than a select group of individuals, especially when the issue entails highly visible and diffuse costs such as military action. Scholars have noted how the public are wary of the national interest being hijacked by corrupt elites and wish to keep in office “good, public-regarding executives” (Downs and Rocke Reference Downs and Rocke1994, 364; Lowi Reference Lowi1964), and how people are generally averse to situations where only a select few benefit: individuals are less likely to accept unequal divisions of goods because they believe such situations to be immoral and unfair (Fehr and Schmidt Reference Fehr and Schmidt1999; Gottfried and Trager Reference Gottfried and Trager2016), and more likely to prefer foreign policies which minimize inequality or benefit the poor over the rich (Herrmann, Tetlock, and Diascro Reference Herrmann, Tetlock and Diascro2001; Lü, Scheve, and Slaughter Reference Lü, Scheve and Slaughter2012). Therefore, based on both self-interest and normative reasons, I predict:

Economic Benefits and Plausibility of Benefit Diffuseness

In practice, however, it is difficult for individuals to assess the exact diffuseness of an international issue’s benefits. After all, how the benefits are distributed is not an inherent quality of the issue but depends heavily on the political institution, domestic politics, and the specific situation of the issue (Cowen Reference Cowen1985; Gowa Reference Gowa1988).Footnote ³ In fact, even within the same issue area of territory, which is usually regarded as benefiting the entire nation (Wright and Diehl Reference Wright and Diehl2016, 34), there exist cases where some are seen as providing concentrated benefits to certain domestic ethnic or political groups. For example, in 1812, the U.S. South vigorously opposed the incorporation of Quebec into U.S. territory because they feared that the addition of Canadian territory would only empower the North at their expense (Maass Reference Maass2020). Similarly, Somalia did not attempt to claim lands inhabited by their co-ethnics on the Ethiopian side of the border when the Majertain family held power because the Majertains feared that the incorporation of Ethiopian-Somalis would only provide more representation to their competitors, the Ogadeni clan, and weaken their political influence (Saideman Reference Saideman1998).

Further complicating the measurement of issue diffuseness is that even the same issue can be viewed as having either diffuse or concentrated benefits depending on which aspect of the issue is emphasized. Those who support fighting for a given good would stress how the good can provide diffuse benefits to all nation members, while the opposition would emphasize how the issue may end up serving only the interest of select domestic groups. This leads to a second question: what kind of issues are more likely to be perceived as providing diffuse benefits to all nation members when faced with conflicting information about who benefits?

I argue that issues with specific economic benefits are less likely to be perceived as providing diffuse benefits to the nation. To be clear, economic benefits do not have to be concentrated; they can certainly provide diffuse benefits to the nation. However, their easier divisibility makes them more easily imagined as excludable.Footnote ⁴ Additionally, because economic benefits are often more specific and tangible, it is easier to associate the benefits with a relevant domestic group who may especially gain. This in turn lends more plausibility to the accusation that the issue exclusively benefits concentrated interests, even when it is not necessarily true. For example, the U.S. President Donald Trump continuously took advantage of the fact that some export industries gain more from lower tariffs to accuse free trade agreements of serving parochial interests. In a well-received presidential rally in Pennsylvania, he declared that “the inner cities will remain poor. The factories will remain closed. The borders will remain open. The special interests will remain firmly in control” (Time, June 28, 2016).Footnote ⁵

Moreover, the ease of identifying domestic groups who can especially benefit from the issue makes it easier for the opposition to introduce concerns of parochial interest involvement. Hobson (Reference Hobson1902, 47), for example, points to the existence of “special commercial and social interests which stand to gain” as evidence for his argument that Great Britain’s national interests had been hijacked by parochial groups. Other scholars also note that the more obvious the private interest at stake, the more leaders would have to try to legitimize their case for pursuing a specific policy (Goddard and Krebs Reference Goddard and Krebs2015).

A good example of how economic benefits are hard to frame as diffuse is the case of the Persian Gulf War in 1991. During the war, the U.S. Secretary of State James Baker claimed that the U.S. intervention in the Gulf was a “vital national interest” because “American jobs” and “the pocketbook and standard of living of every American” were at stake (PBS, November 9, 1990).Footnote ⁶ Although Baker attempted to frame the Gulf War as bringing diffuse economic benefits to the American public, his statements did more to invite suspicion that there were ulterior motives to the war. Hackett and Zhao (Reference Hackett and Zhao1994) find that accusing the war of serving the interest of oil companies was one of the most frequent arguments used by anti-war groups, taking up 33% of reported anti-war newspaper headlines during the war. On the other hand, more abstract values such as benefits of national security or prestige are less subject to accusations of concentration since it is harder to imagine how such issues could only benefit a specific group of people. In addition, because what is exactly at stake is ambiguous, it is also harder for critics to disprove the claim that abstract benefits would be widely distributed. Therefore,

Economic Benefits and Plausibility of Diffuse Non-Economic Benefits

Yet $ {H}_2 $ is still insufficient to conclude that issues with specific economic value would be considered less important to the nation than issues without any economic value. After all, if abstract non-economic benefits were more resilient to claims of concentration, proponents would try to frame all issues, including issues that have economic value, as matters of diffuse non-economic benefits. In fact, politicians have often tried to frame the aforementioned trade issues, wars, and imperialist foreign policies as matters of national security and status rather than matters of economic gains.

This leads to the final part of my argument, where I argue that having economic benefits can make an issue harder to be considered national interest by making it possible for the opposition to raise criticisms of concentrated economic benefits. Even when proponents try to describe the issue as a matter of national security, claims of concentrated economic benefits can provide a plausible alternative way of understanding the issue, undermining the credibility of other abstract benefits involved.

For example, when Bolivia went to war against Paraguay over the Chaco region, a land with rich oil and gas potential, the Bolivian president Daniel Salamanca never mentioned the potential economic benefits from oil as a reason for why Bolivia should go to war (Meierding Reference Meierding2020, 88).Footnote ⁷ Instead, he claimed that gaining the Chaco would benefit all Bolivians by providing a passage to the Atlantic and that the problem of the Chaco was a matter of national honor (Arze Quiroga Reference Arze Quiroga1951). However, Chaco’s rich oil potential provided Salamanca’s political opponents with an opportunity to accuse Salamanca of going to war for concentrated economic interests. The political opposition argued that the real cause of the war was to benefit oil companies and corrupt Bolivian elites who were trying to benefit at the cost of ordinary Bolivians (Marof Reference Marof1935). Despite Salamanca never mentioning oil benefits as a reason for war, and despite ample evidence that oil interests were never actually involved, the Bolivian public “almost immediately accepted as truth that the Chaco War was the result of a basic conflict of oil lands” (Klein Reference Klein1964, 175).

Similarly, just before the 2003 Invasion of Iraq, U.S. Defense Secretary Donald Rumsfeld insisted multiple times that “The conflict with Iraq is about weapons of mass destruction (…) It has nothing to do with oil, literally nothing to do with oil.”Footnote ⁸ Records from classified internal documents also revealed that many government officials, including the U.S. president George W. Bush, were very cautious on the matter of oil and agreed multiple times on the principle that “Iraqi petroleum belonged to the people of Iraq” (Meierding Reference Meierding2020, 161). However, despite the precautions taken and strong denials by the Bush government, the war was met with much public scrutiny about parochial interest involvement. Again, during the Chile–Bolivia war over the nitrate-rich land of Antofagasta in 1879, the role of mining companies was continuously suspected for over a century, even when the Chilean president Anibal Pinto explicitly insisted that “Chile did not go to war to save the Compania de Salitres but to enforce its [sovereignty] treaty rights [with Bolivia]” (Sater Reference Sater1986, 13).

By contrast, accusations of concentrated economic benefits cannot be made when there is no economic value for anyone to benefit from. Consequently, when issues without economic value are framed as matters of national security and honor, they are more likely to remain framed as matters of abstract benefits because alternative framings of economic interests cannot be offered.Footnote ⁹ Even when the issues’ non-economic benefits are unclear, it is relatively easy to maintain that the issues have national value because the existence of abstract benefits is difficult to disprove. For instance, the Siachen Glaciers are not valuable for economic or security reasons to the eyes of outside observers (Chari, Cheema, and Cohen Reference Chari, Cheema and Cohen2003, 53), but because accusations of concentrated economic benefits cannot be offered while the claim that the glaciers could provide some strategic advantage in the future is hard to disprove, the glaciers remain perceived as matters of vital national security for citizens in both India and Pakistan. To summarize, having economic value makes possible accusations of concentrated interests, which are effective in undermining even the sincerity of proponents’ claims that there are other, diffuse abstract benefits to the issue. Consequently,

Moderators and Scope Conditions

Finally, I wish to clarify two key scope conditions of the theory. First, the theory assumes the existence of a visible domestic opposition that is strong enough to offer suggestions of concentration. It is, therefore, possible that the theory is less applicable in authoritarian regimes where such opposition does not exist. To what extent it is less applicable remains an empirical question: after all, effective domestic political opposition can also exist in many non-democracies, where criticisms can be voiced by different factions within a governing party or unofficial civilian opposition outlets. In fact, the cases of Bolivia and Chile referenced above were both semi-authoritarian at the time. Studies also suggest that authoritarian leaders, especially non-personalist ones, are not completely insulated from public opinion and have incentives to care about maintaining public support and legitimacy (Li and Chen Reference Li and Chen2021; Weeks Reference Weeks2012; Weiss Reference Weiss2013).

Second, the applicability of the theory will depend on how excludable the economic benefit is seen to be. Accusations of concentration will be more successful when there is more suspicion about the benefit’s potential to provide concentrated benefits: for example, when the benefit at stake is more easily captured by oligopolistic groups—such as a capital-intensive resource like oil—or when there is higher pre-existing distrust among the public against the redistribution system, the ruling elites, or potential beneficiaries. The empirical section below focuses on testing the three main hypotheses. In the interest of space, I discuss these ideas further in the conclusion and present several related findings in Appendix A.6 of the Supplementary Material.

Study 1: Benefit Diffuseness and Perception as National Interest ( $ {H}_1 $ )

Study 1 examines $ {H}_1 $ , investigating the effect of benefit diffuseness on how likely the issue is perceived as a national interest worth fighting for. Using the survey platform Lucid, I conducted a survey experiment on a nationally representative sample of 1,196 U.S. adults in October 2020. Lucid ensured that the participants resembled the gender, age, geographic, and racial distribution of the U.S. adult population, and the full survey data can be found in the APSR Dataverse.Footnote ¹¹ All survey participants started by reading:

Participants were then introduced to the benefits of taking Seal Island. The survey randomized two variables: the breadth of the population receiving the benefit (Benefit Diffuseness), and the issue area of the benefit (Benefit Type). Benefit Diffuseness varied on three levels, ranging from the most diffuse condition of “All Americans,” to the moderately diffuse condition of “Alaskan Residents,” and to the most concentrated condition of “Employees of Occidental Oil.” Occidental Oil was described as an “American oil company operating in the region.” As for benefit type, respondents read about either a specific economic benefit or a vague security benefit. Respondents assigned to the economic condition were told that taking the island would lead to $150 billion worth of benefits for either all Americans, Alaskan residents, or employees of Occidental Oil. Those in the security condition were told that taking control of Seal Island would lead to increased protection for one of the three beneficiaries. This resulted in a $ 3\times 2 $ variation, where for each benefit type the prompt read:

Those who did not receive “All Americans” as their beneficiary condition were explicitly told at the end of the prompt that “Other Americans, apart from [Alaskan residents/employees of Occidental Oil] are not expected to benefit from taking Seal Island.” Respondents were then asked a series of questions assessing their opinions of Seal Island: how much they approved or disapproved of using the U.S. military to take control of Seal Island (Support for Military Operation);Footnote ¹² how much they agreed or disagreed with statements related to national interest, such as “Taking control of Seal Island is in the U.S. national interest” (National Interest), and “Taking control of Seal Island benefits the U.S.” (Benefit U.S.); and statements related to respondents’ level of personal engagement with Seal Island, such as “Taking Seal Island would make me proud” (Proud to Have), “I would be unhappy if Russia took control of Seal Island” (Unhappy to Lose), and how willing participants were to contribute $20 to fund a military operation for Seal Island should the U.S. government request it (Contribute Money).Footnote ¹³ The full prompt and questionnaire are provided in Appendix D of the Supplementary Material.

Each point in Figure 2 represents the mean value of each outcome variable depending on the level of benefit diffuseness. Figure 2a displays results for when the benefit at stake was security, and Figure 2b shows results for when the benefit was economic. In both figures, perception as national interest, level of support for military action, and level of personal engagement were the highest when all Americans benefited, followed by when only Alaskans benefited, and the lowest when only employees of Occidental Oil benefited. The differences in the mean values for the three diffuseness conditions were significant at $ p<0.05 $ for all six dependent variables (Appendix A.1.4 of the Supplementary Material).Footnote ¹⁴ This pattern is consistent across both benefit types, indicating that both security and economic benefits are less favored when they are concentrated. The results support $ {H}_{1a} $ through $ {H}_{1c} $ , which predicted that the more diffuse the issue’s benefits, the more likely the issue would be thought of as a national interest and worth fighting for.

Figure 2. Effect of Benefit Diffuseness on National Interest

Note: All dependent variables were measured on a five-point scale except for Contribute Money which was measured on a four-point scale. Each point indicates the mean value of each dependent variable according to the beneficiary condition received. The differences in mean values by beneficiary condition are significant at $ p<0.05 $ . It is also worth noting that economic benefits are not always considered less important than security benefits when controlling for their level of diffuseness. For the related regression, see Appendix A.1.1 of the Supplementary Material.

Substantively, changing beneficiaries from all Americans to Alaskan residents or employees of Occidental Oil each resulted in about a 12% and 20% decline in the level of support for all dependent variables.Footnote ¹⁵ I also find that the decline is due to both self-interest and normative reasons as suggested earlier in the article: the more concentrated the benefits, the less likely respondents believed that they would personally benefit from the issue or that Americans as a whole would benefit from it. These beliefs in turn significantly lowered respondents’ support for the issue as a national interest worth fighting for—detailed results using causal mediation and examples of open-ended responses are available in Appendix A.2 of the Supplementary Material.Footnote ¹⁶ Preference for diffuse benefits was also robust to controlling for participants’ demographic and attitudinal qualities such as party identification, income, education, gender, militantism, nationalism, and friendliness toward Russia (Appendix A.1.2 of the Supplementary Material), and the treatment effects were remarkably homogeneous across respondents’ party identification and political ideology (Appendix A.3 of the Supplementary Material).

In addition to examining changes in mean values, I also analyze whether there is more convergence in opinions when benefits are diffuse (Table 1). After all, a highly divisive issue cannot be seen as a national interest since such issues would have a harder time eliciting broad public support for conflict escalation. Table 1 presents the variance of each dependent variable depending on the level of benefit diffuseness. A smaller variance in responses implies that opinions are more converged, while a larger variance indicates that opinions are more dispersed (Kertzer, Rathbun, and Rathbun Reference Kertzer, Rathbun and Rathbun2020). We observe that variances in the answers are the lowest when all Americans benefit and the highest when only Occidental Oil employees benefit for most of the dependent variables.Footnote ¹⁷ The differences in variances are also significantly different from each other for the most part, as shown in the final column. These results indicate that issues with concentrated benefits not only receive lower support on average, but also induce more divergence in beliefs about the issue’s national importance.

Table 1. Convergence in National Interest Perceptions by Benefit Diffuseness

Note: Values indicate the variance in the responses for each DV per Beneficiary condition. Lower numbers indicate more convergence in answers. The final column indicates whether the variances per Beneficiary condition are significantly different from each other: the p-values indicate how confidently we can reject the null hypothesis that the three variances are equal.

Study 2: Economic Benefits and Vulnerability to Suggestions of Concentration ( $ {H}_2 $ )

Study 1 showed that how an issue’s benefits are distributed domestically matters in deciding whether people perceive the issue to be important to the nation and worth fighting for. Study 2 builds on these findings to examine whether economic benefits are more vulnerable to the accusation of providing concentrated benefits compared to abstract, non-economic benefits. The second experiment was carried out on a nationally representative sample of 825 U.S. adults in March 2021. As in Study 1, participants read about a hypothetical dispute over Seal Island and the anticipated costs of taking the island. Depending on which benefit type condition participants were assigned to, participants then either read about an economic benefit or an abstract security benefit for all Americans.

The innovation in Study 2 is that it introduces uncertainty about who benefits. Instead of stating as a fact that taking the island would benefit a certain group of beneficiaries, all participants were suggestively told that “Some policymakers argue taking Seal Island would benefit all Americans.” Half of the participants then received an additional message that other policymakers believed taking the island would only benefit certain interest groups:

This design yields a $ 2\times 2 $ variation, where the conditions were randomized across whether participants read about the economic or security benefit of taking the island (Benefit Type), and whether participants read about a possible concentration of benefits (Concentration Suggested). This design allows us to compare how effective the opposition’s accusation of benefit concentration is depending on the issue’s Benefit Type. The main dependent variable in this survey is how much the issue is seen as providing diffuse benefits to all Americans, measured as the level of agreement with the statement “Taking control of Seal Island benefits the American people” (Benefit Americans) on a five-point scale. If $ {H}_2 $ holds, reading about the opposition’s accusation would lower the average level of agreement with Benefit Americans more for the economic benefit condition than for the security benefit condition.

Table 2 displays the four possible treatment conditions and the average level of agreement with Benefit Americans. I also present results for how nationally important people perceived the issue to be, measured as the level of agreement with the statements “Taking control of Seal Island benefits the U.S.” (Benefit U.S.) and “Taking control of Seal Island is in the U.S. national interest” (U.S. National Interest). The results are also consistent across other outcome variables measuring levels of personal engagement (Proud to Have, Unhappy to Lose, and Contribute Money) and level of support for military operation (Support Military Use), presented in Appendix B.3 of the Supplementary Material. All the dependent variables were measured on a five-point scale,Footnote ¹⁸ and the effect of receiving the opposition rhetoric is summarized in the second to last row with its 95% confidence intervals.

Table 2. Persuasiveness of Concentrated Benefits Suggestion by Benefit Type

Note: The table shows the average level of each dependent variable for each treatment condition (A-D). People assigned to conditions A and B only read about the possible diffuse benefits of taking Seal Island, while those assigned to C and D also read about how the benefits could be concentrated on certain groups. “Suggestion Effect” calculates the difference-in-means of the DVs between A&C and B&D with 95% confidence intervals. The last row displays the relative effect size of the concentration suggestion in the strategic locations condition compared to the oil fields condition. Asterisks indicate whether the difference in the effect size is statistically significant. * $ p<0.1 $ , ** $ p<0.05 $ .

Table 2 demonstrates that suggestions of benefit concentration reduced the level of agreement with Benefit Americans by 0.78 points in the oil fields condition, compared to only 0.43 points in the strategic locations condition. As can be seen from the last row of the table, the opposition was also almost twice as effective in undermining people’s beliefs that the issue would benefit the United States (Benefit U.S.) or be in the national interest (National Interest) when the benefit was economic compared to when it was more abstract. The heterogeneous effect of suggesting benefit concentration by benefit type is visualized in Figure 3.

Figure 3. Persuasiveness of Concentrated Benefits Suggestion by Benefit Type (Visualized)

Note: The graphs visualize Table 2 results, the effect of reading about a possible concentration of benefits depending on Benefit Type. The dependent variables are written on the y-axis and range from Strongly Disagree (1) to Strongly Agree (5). Robust confidence intervals are shown at 95%. Regression results of the interaction between Benefit Type and Concentration Suggestion are in Appendix B.2 of the Supplementary Material; the interaction is significant at $ p<0.05 $ for Benefit Americans and Benefit U.S.

Furthermore, I find that suggestions of benefit concentration are also more effective at dispersing beliefs about who benefits in the economic condition than in the security condition. Figure 4 shows the density distribution of responses for Benefit Americans when the benefit is strategic locations (left) and when the benefit is oil fields (right). Responses from participants who did not receive the opposition’s argument are depicted in light gray, while responses from participants who read about a possible concentration are shown in darker gray. In the case of security benefits, although respondents were generally less likely to agree that taking Seal Island would benefit Americans when introduced to the possibility of concentrated benefits, the shape of the distribution did not change dramatically. In fact, the proportion of participants who either strongly or somewhat agreed with the statement declined by only 18 percentage points, from 69% to 51% (Figure 4a). In contrast, when oil fields were at stake, reading about the possible concentration of benefits reduced the proportion of those who strongly or somewhat agreed with the statement by 30 percentage points, from 73% to 43% (Figure 4b). Those who strongly or somewhat disagreed that taking Seal Island would benefit the American people also increased by almost three times in the oil fields condition, from 12% to 32%.

Figure 4. Persuasiveness of Concentrated Benefits Suggestion by Benefit Type (Belief Dispersion)

Note: Smoothed density plot showing the distribution of responses for “Taking Seal Island benefits the American people” (Benefit Americans). The responses range from Strongly Disagree (1) to Strongly Agree (5) and are plotted on the x-axis. The difference in variance is only significant for panel (b), the economic condition (Appendix B.1 of the Supplementary Material).

Finally, akin to Study 1, economic benefits were not necessarily less preferred to security benefits when the opposition did not cast doubt on how the benefits would be distributed. In other words, while economic benefits were more vulnerable to accusations of concentrated benefits as hypothesized by $ {H}_2 $ , people did not automatically associate economic benefits with concentrated interests in the absence of such accusations. This finding has two important implications. One, economic benefits are not inherently less preferred than security benefits when their distributional consequences are perceived as diffuse. Two, it highlights the theory’s scope conditions: the theory may be less applicable to situations or polities where the political opposition does not or cannot raise accusations of benefit concentration.

Study 3: Potentially Concentrated Economic Benefits versus No Benefits ( $ {H}_3 $ )

Studies 1 and 2 have each shown that the public prefers diffuse benefits over concentrated benefits and that economic benefits are more easily discredited as concentrated than non-economic benefits. The third survey tests the final hypothesis ( $ {H}_3 $ ), which predicted that issues suggested of having concentrated economic benefits would be less likely to be considered national interest than issues without any economic benefits. Study 3 was fielded on February 12–14, 2022, on 677 U.S. adults.

Study 3 employed a simple design, varying only whether or not the issue at stake had objective economic value that could be accused of concentrated economic benefits. When introducing a hypothetical territorial dispute over Seal Island between the U.S. and Russia, half of the respondents were told that “the island does not have any natural resources” (No Value), while the other half were told “the island provides access to offshore oil fields” (Economic Value). After being told of the anticipated costs of taking over the island, all respondents in both the No Value and the Economic Value scenarios read about the proponents’ attempt to frame Seal Island as a matter of diffuse security benefits:

Next, all respondents were introduced to opposing arguments. In the case where the island had no value, the opposition cast doubt on the size of Seal Island’s security benefits by saying, “They claim the security benefits provided by Seal Island would not be as large as what some policymakers argue.” In the case where oil was present, the opposition was able to make an additional accusation about how the oil benefits would be concentrated on certain domestic groups, which read, “In fact, they say the real reason for trying to take Seal Island is to benefit certain oil companies, not to protect all Americans. Only some oil companies would benefit from taking the island, at the cost of other Americans.” This design allows us to test how effective an accusation of concentrated economic benefits is in decreasing support for an issue that is objectively more valuable relative to when the issue does not have any economic value and is not subject to such accusation.

After reading the prompt, respondents were asked how much they approved or disapproved of a military operation to take Seal Island (Support Military Use), followed by questions on whether taking Seal Island would be beneficial for most Americans (Benefit Americans), for the U.S. (Benefit U.S.), or be in the national interest (National Interest), presented in random order. Participants were also asked how willing they were to contribute $20 to support a military operation for the island should the U.S. government request it (Contribute Money).

The results in Table 3 show respondents consistently believed that gaining an island with potentially concentrated economic benefits would result in fewer benefits for Americans and the nation compared to when the island did not have any economic value. For example, the belief that taking Seal Island would benefit Americans was lower by 19.5% when the island had economic value with uncertain distributional consequences compared to when the island did not have any clear value to begin with (model 2). Respondents were also less likely to support military use over Seal Island or to contribute personally to the dispute in the economic value condition.

Table 3. Effect of Having Economic Value with Uncertain Distributional Consequences

Note: Coefficients show how the DVs change relative to when the island does not have any economic value. Coefficients are estimated using ordinary least squares and robust SEs are presented in parentheses. All responses were measured on a five-point scale; support for military use is significant at $ p<0.05 $ when using the seven-point scale. Full regression results with covariates are available in Appendix C.2.2 of the Supplementary Material. * $ p<0.1 $ , ** $ p<0.05 $ , *** $ p<0.01 $ .

Finally, I investigate Table 3 results in more detail by examining the theoretical mechanism suggested by $ {H}_3 $ . $ {H}_3 $ predicted that having economic value can become a liability in framing an issue as a national interest because it opens up the issue to suggestions of concentrated economic benefits, which in turn can cast doubt on the sincerity of the issue’s other non-economic benefits. To test this mechanism, I first examine whether respondents were indeed less likely to believe that the island had diffuse security benefits when the island had economic value. Figure 5a displays respondents’ level of agreement with the statement, “The policymakers who support taking Seal Island are trying to take the island for national security reasons” (Matter of National Security), both for when the island had no economic value (left) and for when the island had economic value suggested of concentration (right).Footnote ¹⁹ The figure shows that when the island had potentially concentrated economic value, less than half of the respondents (45%) either agreed or strongly agreed with the proponents’ claims that taking Seal Island was a matter of national security. By contrast, when the island had no economic value, 75% of participants believed policymakers were pursuing the issue for national security reasons even when they had read about the island’s doubtful security benefits.

Figure 5. Testing $ {H}_3 $ Mechanisms

Note: Panel (a) shows the distribution of responses for Matter of National Security measured from Strongly Disagree (1) to Strongly Agree (5). Box plots display median and quartile values. The mean values for the two conditions are 3.98 (No Value) and 3.10 (Economic Value), and the difference is significant at $ p<0.01 $ . Panel (b) displays the regression coefficients of Matter of National Security for each dependent variable presented in Table 3. Confidence intervals are at 95% and full mediation results are available in Appendix C.1 of the Supplementary Material.

I then examine the second step of the mechanism—whether belief in the issue as a matter of national security affected respondents’ view of the issue as a national interest worth fighting for. To do so, I estimate the effect of Matter of National Security on each of the dependent variables presented in Table 3. Because respondents’ perception of the issue as a matter of national security is not randomized, in this estimation stage I control for other covariates that can affect its relationship with the dependent variables, such as party, education, gender, and income. The results are displayed in Figure 5b and demonstrate a strong positive relationship between respondents’ belief of Seal Island as a matter of national security and their level of support for the five dependent variables.

Overall, Study 3 demonstrates that issues with economic benefits, when faced with the suggestion that the benefits may be concentrated, are less likely to be seen as being in the national interest than issues without any economic value because the economic benefits also undermine the sincerity of non-economic benefits. By showing how the presence of economic benefits can backfire, this result challenges the conventional wisdom that economic benefits make an issue more valuable all else equal. It also provides a potential explanation for some puzzling empirical findings in the territorial conflicts literature, where studies have found that contested lands with resources are less likely to be claimed or evolve into militarized conflict (Altman Reference Altman2020; Hensel and Mitchell Reference Hensel and Mitchell2005; Huth Reference Huth1996; Lee Reference Lee2023b).