The United States (U.S.) commands one of the world’s largest and most expensive advanced militaries with an ever-increasing global presence. As of September 2022, there are 171,736 active-duty military troops across 178 countries including approximately 750 U.S. military bases in at least 80 countries, with Japan, Germany, and South Korea having the largest number of bases – 120, 119, and 73, respectively (Allen et al. 2022). These host countries serve as key military and political allies and economic partners of the U.S., and their cooperation is deemed vital for both American and global security.

Overseas U.S. military bases, however, have become a subject of controversy in many host countries. The controversies are multifaceted, ranging from value discrepancies between Americans and the host country public to concerns about crimes committed by U.S. soldiers, with these scandals undermining public attitudes towards U.S. bases and the U.S., potentially straining diplomatic relations between the U.S. and the host country (Allen and Flynn 2013; Kim et al. 2022; Kim and Boas 2019; Shinomoto 2023; Yeo 2011). This underscores the importance of the public diplomacy functions of a foreign-based military.

While existing literature seeks to conceptualize and describe U.S. military public diplomacy (Karadag 2016; Wallin, 2015; Williams 2009), few have analyzed the effectiveness of U.S. military public diplomacy through online channels and in allied countries. This paper aims to fill this gap by investigating the effectiveness of the U.S. military’s digital public diplomacy on social media, targeting the allied countries hosting the U.S. military bases. Considering the findings from existing studies that public opinion about U.S. foreign policy in foreign countries does affect their policies toward the U.S. (Goldsmith and Horiuchi 2012), it is of vital interest for the U.S. to maintain positive perceptions from the host country’s public towards the U.S. military. Maintaining the host country’s government’s desire for a continued alliance is partially dependent on citizens’ attitudes toward the alliance.

Thus, our research question is: Does the U.S. military’s digital public diplomacy enhance the image of the U.S. military abroad, and consequently, the perception of the U.S., and if so, what types of messaging are more impactful?

Departing from existing conceptual studies on military public diplomacy, we categorize U.S. military public diplomacy in the digital sphere into two groups: one highlighting the traditional military functions of the U.S. military, and another centering on its non-traditional roles. Beyond their primary role in serving the goals of military alliances, the U.S. military overseas is increasingly active in directly engaging with citizens of the host countries to win their hearts and minds (Boone 2021; Wallin 2015; Williams 2009). Drawing from theoretical and empirical studies on the effectiveness of public diplomacy as well as social psychology research, we posit that while both categories of messages are expected to enhance the perception of the U.S. military in host countries, the latter, emphasizing non-traditional roles, is anticipated to have stronger, more positive effects.

We examine our theoretical expectations by concentrating on the U.S. military deployments in South Korea. As a longstanding military ally of the U.S. since the Korean War, South Korea hosts one of the largest contingents of U.S. bases and active-duty military personnel, collectively known as the United States Forces Korea (USFK) (Allen et al. 2022). While the population has generally been supportive of the alliance with the U.S., the country has also witnessed significant anti-base sentiments among the population, driven by nationalist sentiments, fears of looming regional conflict, and NIMBY sentiments (Allen and Flynn 2013; Allen et al. 2020; Kim and Boas 2019; Kim et al. 2022). This context positions South Korea as a prime target for integrating new military public diplomacy strategies into U.S. troop deployment.

To test our expectations, we designed and implemented an original survey with a sample of 626 South Korean citizens in October 2023. Based on the content analysis of 177 messages posted by the USFK’s official Twitter account, @USForcesKorea, we created two faux tweets, labeled “traditional messaging” and “non-traditional messaging,” respectively. We used these faux tweets as information treatments in our survey-embedded experiments. We find that U.S. military messaging has a significant impact on improving perceptions of the USFK and the U.S. in South Korea. Both traditional and non-traditional messaging approaches proved effective in influencing public perceptions. Differences in the effect sizes between these messaging types were not substantial. One unexpected additional finding was that the non-traditional treatment is statistically significantly and negatively associated with fears of imminent conflict on the Korean Peninsula. Altogether, this research provides compelling evidence that U.S. military messaging on social media can have a positive impact on changing perceptions of both the U.S. military and the U.S. in allied countries. At the same time, our findings highlight the need for future studies that further investigate the nuanced effects of different messaging approaches.

The remainder of the paper is structured as follows: we begin by providing an overview of relevant existing literature, including conceptual studies on military public diplomacy and research on digital public diplomacy by the military. Next, we present our theoretical argument in light of this literature. Our research design is then discussed, followed by a report on our main findings. We conclude with a discussion of the implications of our findings and suggestions for future research avenues.

Over the past decade, there has been a growing interest in public diplomacy among both scholars and practitioners. Effective public diplomacy creates a positive image of the engaging country among foreign publics (Goldsmith and Horiuchi 2009; Goldsmith et al. 2021). This positive perception is then expected to influence the actions of foreign governments in ways that benefit the engaging country (Malone 1985; Datta 2009; Goldsmith and Horiuchi 2012; Plouffe and Slingsby 2019; Ahn and Lim 2023). Various actors have been recognized as key players in public diplomacy, including political leaders (Goldsmith et al. 2021), embassies (Yun and Lim 2022), national cultural institutes (Hartig 2012), and official aid agencies (Blair et al. 2022; Kim and Lim 2023). The role of the military in public diplomacy has also garnered attention.

In the remainder of this section, we review the literature on military public diplomacy, establishing the groundwork for our study of USFK as a key actor in public diplomacy. We then narrow our focus to studies examining the use and effectiveness of military public diplomacy in the digital space, identifying gaps in the literature. This discussion paves the way for the theoretical arguments presented in the next section.

Military Public Diplomacy

Karadag (2016), in his conceptual work on military public diplomacy, proposes three tools that military can employ in its public diplomacy: personnel training, humanitarian and disaster relief (HADR) missions, and territorial military support. This section reviews existing empirical studies pertaining to each tool. The majority of the literature centers on military public diplomacy conducted by the U.S., with Israel being the second most frequently studied country.

First, in the domain of personnel training, the U.S. offers the International Military Education & Training (IMET) service to foreign militaries, aiming to provide training and strengthen alliances with the U.S. Several existing studies have explored the effectiveness of such training programs. For instance, Martinez Machain (2020) suggests that nations with a higher proportion of their military personnel trained in the United States are more likely to align their foreign policy preferences with those of the U.S. She argues that this alignment may result from the trainees adopting a mindset that naturally aligns with U.S. preferences. Alternatively, the U.S. might gain influence with military officers within the political elite, facilitating persuasion for the execution of more pro-U.S. policies.

In exploring the second tool, humanitarian assistance and disaster relief (HADR), Flynn et al. (2019) investigate the U.S. military’s HADR missions in Peru. They underscore the pivotal role played by the U.S. military in delivering foreign aid after disasters, encompassing activities such as reconstructing facilities, administering immunizations, and promoting educational and agricultural development. The study reveals that direct engagement with U.S. military personnel correlates with more favorable attitudes toward the U.S. government and its military. The authors, at the same time, highlight that short-term deployments might not fully harness the potential to enhance positive perceptions of the U.S. Their findings suggest that long-term deployments, exemplified by active foreign military bases, could be more effective in cultivating positive public opinion from civilians compared to brief humanitarian assistance following a natural disaster.

Finally, the last tool, territorial military support, describes the development and maintenance of roads, schools, and other infrastructure in a foreign country by a military outside of immediate disasters, with much of the U.S. territorial military support taking place in developing sub-Saharan Africa and Southeast Asia. In Africa, the U.S. military used its presence to build rehabilitation centers to treat children suffering from malnutrition, ultimately gaining the goodwill of the local people, culminating in successful public diplomacy (Brown 2013).

While Karadag’s (2016) conceptual framework is insightful, the military public diplomacy initiatives reviewed in his framework are primarily carried out in developing target countries (such as humanitarian aid and disaster relief, and territorial military support) or during specific circumstances (such as natural disasters), or on a short-term basis (military training and exercises). It lacks discussion on durable, peacetime military public diplomacy that can be deployed in economically advanced host countries.

Military Public Diplomacy in a Digital Space

While existing studies extensively explore digital public diplomacy practices of major countries through the social media accounts of political leaders and embassies (Mattingly and Sundquist 2023; Mattingly et al. 2024; Dodd and Collins 2017), there is a surprising lack of research on the military’s digital public diplomacy. Studies indicate that the U.S. State Department and the Department of Defense have been slow to integrate the web into their operations. Instead, they have focused primarily on unilateral communication through traditional media (Samuel-Azran and Yarchi 2018; Williams 2009) or physical leaflets (Wallin 2015), whose tactical effectiveness in winning the hearts and minds of the target population has been mixed (Braun 2012).

Empirical research on the use of online channels by the U.S. military is thus limited. One notable study by Boone (2021) examined U.S. operations during the early years of the war on terror. In Iraq and Afghanistan, the U.S. military employed strategic influence operations by uploading videos filmed by soldiers to YouTube. These videos aimed to refute human rights abuse allegations and humanize the troops but were generally perceived as psychological operations and widely regarded as propaganda (Boone 2021). In contrast, Israel used Facebook as a strategic enactment of smart-power, combining hard-power operations with soft power in the form of two-way online communication, attempting to promote the actions of Israel Defense Forces (IDF) positively to the Muslim world (Samuel-Azran and Yarchi 2018). The authors found that the IDF Spokesperson’s Unit effectively engaged the Arab and Muslim world by highlighting shared values and mutual interests between Jews and Arabs, achieving significant interaction through online communication. This approach of targeting specific demographics and maintaining an open online channel proved beneficial for the military during conflict.

As seen from the two studies reviewed above, the discussion on military digital public diplomacy has primarily centered on its use in active conflict scenarios and efforts to garner support from adversarial or third-party audiences. There is a notable lack of research on the military’s use of digital communication during peacetime, particularly in the context of allied countries. This research aims to fill that gap. In doing so, we build on the fundamental notion that security threats and security commitments surrounding an alliance are “not fixed perceptions” but depend on the “dyadic political relationship between states and subsequent interpretations of the changing political relationship” (Son and Yim 2021: 3). For the general public who do not have complete information on the geopolitics and volatile security situations affecting their country, both the perceptions of threat and the legitimacy of protection would depend on how they are communicated. Digital public diplomacy that reaches a broad audience in target countries has the potential to present the positives of the alliance in a way that might outweigh the negatives in the eyes of the public (Merickova 2014; Horiuchi and Tago 2023; Yeo 2011).

In the following section, we discuss the potential messages overseas U.S. military bases can utilize in their digital communication towards the public of the allied countries and derive our hypotheses.

Synthesizing existing studies on the military’s public diplomacy, we categorize the digital communication efforts of the U.S. military into two types based on message content: the first type comprises of messaging of traditional roles focusing on security-related functions, such as assurance for military protection and military personnel training. The second type encompasses messaging not directly related to security. This includes not only humanitarian aid, disaster relief, and community development, primarily carried out in developing host countries, but also broader engagement with the local civil society which also takes place in economically advanced host countries.

We posit that messages communicating both functions have the potential to improve the image of overseas U.S. military bases among host country publics. On one hand, messaging of traditional roles can increase support for the USFK by effectively priming the common threat and the U.S. commitment to protecting allied countries against these threats. Existing studies have indeed shown that perceptions of the need for U.S. military bases vary depending on present security concerns, with support increasing when tensions with adversaries are higher (Kim et al. 2022; Rich et al. 2022). On the other hand, messages highlighting non-traditional functions and emphasizing the military’s peaceful interaction with local civil society can also be effective by mitigating the perceptions of negative externalities and increasing feelings of being respected despite the context of an asymmetric military alliance (Bueno and Martínez 2023; Son and Yim 2021). This leads us to the following main hypothesis.

Hypothesis 1) U.S. military online messaging positively influences public opinion regarding U.S. troop deployments in host countries.

We further expect the effectiveness of non-traditional messaging centering on civil society engagement to be more effective than traditional security-centered messaging. This is for two reasons. First, these non-traditional roles are less known to the public so messages about them can provide a newer perspective on the recipient’s common assumptions. As discussed in the next section, our content analysis of 177 messages posted by the USFK’s official Twitter account, @USForcesKorea, reveals that traditional security-centered messaging (totaling 133) is much more common than non-traditional messaging (totaling 44). Studies in social psychology and social media find that more novel, information is more likely to produce changes in attitudes and opinions than information that is likely to be already known to the audience (Schnabel and Hristov 2010; Vosoughi et al. 2018; Wang et al. 2021).

Second, military functions of the U.S. military might be subject to partisan and ideological bias, whereas messages on non-traditional roles can be constructed to resonate with a broader audience (Gartner and Segura 1998; Tomz et al. 2020). By not advocating for any particular strategic military action or using divisive language that could split audiences ideologically, showing messages highlighting horizontal and peaceful interactions between the U.S. military and civil society could be more effective at garnering bipartisan support (Mattingly and Sundquist 2023). Based on the above discussion, we propose the following second hypothesis.

Hypothesis 2) Non-traditional messaging from the U.S. military is more effective in garnering positive perceptions of U.S. deployments abroad.

Case selection: South Korea

To test our hypotheses, we designed and implemented an original survey in South Korea. Over the years, as the ROK-U.S. alliance has evolved and South Korea has undergone democratization, public support for the alliance in the country has fluctuated (Son and Yim 2021). Despite the persistent belief in the importance of U.S. security guarantees, support for the alliance diminishes with perceptions of inequality in the relationship (Shinomoto 2023; Snyder et al. 2019). Wariness of negative externalities such as noise, pollution, and crime, combined with heightened concerns of attacks, is also palpable. These factors may lead citizens to reject the presence of the alliance in their local community, even when they consider the U.S. military’s presence vital for their country’s national security (Allen and Flynn 2013; Horiuchi and Tago 2023). Such negative perceptions not only tarnish the image of the U.S. military abroad but, if strong enough, could influence host government decision-making, potentially weakening their alliance with the U.S. (Yeo 2011). Given this context, South Korea emerges as an important case and a prime target for integrating new military public diplomacy strategies into U.S. troop deployments.

Survey design

To test our expectations discussed in the previous section, we designed and conducted an original survey with a sample of 626 South Korean citizens in October 2023.¹ The survey was distributed online by Lucid, an international market research firm. To ensure the quality of responses, a pre-treatment attention check question was included following the demographic questions (Aronow et al. 2019). Only respondents who passed the attention check were allowed to proceed with the survey. The survey incorporated an experimental element, exposing respondents to a random information treatment. One-third of the respondents were exposed to a traditional Twitter message from USFK, while another third received non-traditional messaging from the USFK. The remaining third served as the control group and did not receive any messaging.

We chose Twitter (now X)² as our preferred medium, influenced by the growing popularity of Twitter as a data analysis tool and the active Twitter presence of the USFK. Currently, online platforms directly managed by the USFK are few while its Twitter account is regularly updated. More formal online news platforms suffer from a lack of direct management from the USFK, therefore turning instances of public diplomacy into marketing, which is not the focus of this study. Additionally, Twitter offers the capability to reach a much broader international audience compared to the USFK website (Brantner and Pfeffer 2018). Despite this potential, @USForcesKorea’s tweets primarily target an English-speaking audience and are usually only posted in English, resulting in limited reach within South Korea. Very few are written directly in Korean.³ Tweets with substantial interaction are often those retweeted by official South Korean channels, such as the Korea Herald and the U.S. Embassy in Seoul, and translated into Korean. This suggests that there is a reachable South Korean audience through this platform who are seemingly more willing to engage with USFK-made content when it is accessible for them.

To construct the two information treatments, we first undertook content analysis of the official USFK account, @USForcesKorea. The account posts quite frequently, often every day and multiple times a day, leaving the number of analyzable tweets significantly sizable, which was narrowed down by date,⁴ relevance,⁵ and additional determining factors.⁶ From this sample set, using non-mutually exclusive keywords, tweets were filtered into two sets: traditional and non-traditional functions. Using the keywords “protect,” “enemy,” “defense,” and “deterrence,” we categorized tweets emphasizing traditional military functions accompanied by images of weaponry, technological capacity, and capabilities, or statements directly naming countries it defers as threats, such as North Korea and China, compiling a total of 133 posts. For the non-traditional set, we used the keywords “values,” “culture,” “civilians,” and “event,” again with accompanying images that included meetings with local governments, sponsorship of local events with South Korean citizens, engagement in local activities, families, or otherwise non-military interactions with civilians, totaling 44 posts.

After manually analyzing the posts from both categories, we created two faux tweets, labeled “traditional messaging” and “non-traditional messaging.” Both were reflective of the @USForcesKorea posting style, with the text itself made to emulate the tone of previous posts as closely as possible. The traditional message reads: “Alliance Deterrence is Key. We will not be swayed by enemy threats. We are committed to providing a combined defense posture to protect & defend the ROK against any adversary. #ROKUSAlliance #WeGoTogether #CombatReadiness.” It also contains four pictures that were taken directly from the account itself showing weapons, machinery, and the U.S. and Korean forces together in a simulated combat setting. The non-traditional message reads: “Happy Culture Exchange Day! Our service men and women spent the day with a Korean speech competition, culture events, a film festival, K-pop concert, and sports competitions! #ROKUSAlliance #WeCelebrateTogether.” It is shown with four pictures including USFK soldiers speaking Korean and interacting with Korean civilians while wearing traditional Korean clothing, taking pictures together, and playing sports.

Directly following the treatments, participants were asked two questions to directly evaluate their affinity for the USFK and the U.S., serving as our main dependent variables (USFK Favorability, U.S. Favorability). The subsequent four questions were used as additional dependent variables to gauge respondents’ opinions on the USFK’s perceived functions: the first two questions measure perceptions of non-traditional functions (Civil Society, Aligned Values), while the other two assess perceptions of traditional military functions (Protection, Number of Troops). Lastly, one question was used to measure respondents’ conflict perception (Conflict Perception). Table 2 summarizes the measurement scales and question wording for these dependent variables.

The survey also included a set of demographic questions: gender (male : 1 if male, 0 if female), age (1) under 25 years old, (2) 25-34 years old, (3) 35-44 years old, (4) 45-54 years old, (5) 55-64 years old, (6) 65 and above, and residential location at the provincial level. These were pre-treatment questions. Additionally, we included a question on education (university education : 1 if graduated from a four-year university. This question was asked towards the end of the survey, after the treatment and dependent variable questions. Descriptive statistics are presented in the appendix.⁹

This section reports our main findings. Before presenting the effect of our information treatments, we conducted a balance test and report the results in the Appendix (Table A2). We find that the group that received the non-traditional messaging has a higher proportion of men compared to the control group, with this difference in means being statistically significant (p<0.05). Later in our regression analysis, we control for gender and also explore heterogeneous treatment effect by gender.

Main Findings

The impact of our treatment can be assessed by comparing respondents’ perceptions of USFK and the U.S. across different groups. Figure 1 presents the averages of USFK Favorability and U.S. Favorability using a 1-5 Likert scale ranging from very unfavorable to very favorable. Error bars indicate a 95% confidence interval. The “Control” group comprises individuals who did not receive any message, while the “Traditional” group includes those exposed to messages highlighting USFK’s traditional role. The “Non-traditional” group received messages centered on non-traditional roles. On average, both treatment groups rated the USFK and the U.S. more favorably compared to the control group, consistent with Hypothesis 1. However, we do not find significant differences between the two treatment groups, as indicated by overlapping confidence intervals. We thus do not find support for Hypothesis 2.

Although our experiment is based on the randomized assignment of treatments, we also conducted a series of multivariate regression analyses controlling for pre-treatment demographic variables as well as university education to reduce the variance within experimental conditions and more precisely estimate the effect of the treatments (Mutz et al. 2019: 36). The results are summarized in Table 3. For ease of interpretation, we present the results from linear regression in the main text, with ordered logistic regression results provided in the Appendix (Table A3). Table A4 in the Appendix additionally reports results from employing a 6-category age variable and a 3-category education variable as factor variables (with the baseline being under 25 for age and high school or less for education).

In Models 1 and 5, both Traditional and Non-traditional messaging have statistically significant and positive coefficient estimates, indicating their positive influence on USFK Favorability (Model 1) and U.S. Favorability (Model 5). Regarding the control variables, being male is strongly associated with both U.S. and USFK Favorability. Age has a weakly negative association with U.S. Favorability. Table A4 in the Appendix, where age is entered as factor variables, show that those in the middle age groups (45-54 and, to a lesser extent, 55-64) are less favorable toward USFK and the US compared to those under 25. In contrast, university education does not exhibit any consistent or significant association with USFK or U.S. Favorability.¹⁰

We also explore heterogeneous treatment effects by respondent characteristics. Models 2 and 6 interact our information treatments with gender (Male). We visualize the interaction effect from Model 2 in terms of predicted favorability with 90% prediction intervals in the Appendix (Figure A1). Among women, traditional messaging is found to be more effective than non-traditional messaging. Conversely, among men, non-traditional messaging proves to be more effective. We speculate that these differences may arise because the familiarity with the type of messaging varies by gender. For women, traditional security messaging might provide new information, while men, who may be more familiar with the security functions of the USFK due to mandatory military service, might find non-traditional messaging relatively more novel. Models 3 and 7 interact the treatments with age. The interaction effect from Model 3 is also visualized in the Appendix (Figure A2). Both treatment effects are found to be more pronounced among older age groups. Models 4 and 8 interact the treatments with university education. The predicted values plot in the Appendix based on Model 4 (Figure A3) suggests that, while the differences between university graduates and non-graduates are not statistically significant, the treatment effects are primarily driven by university graduates, who constitute 65% of the sample population.

Additional analysis

In addition to the previously examined dependent variables, our survey explored respondents’ sentiments regarding four specific statements concerning the USFK and the U.S. (see Table 2). The statements aimed to capture perceptions of four distinct aspects of the USFK: 1) Civil Society, reflecting non-traditional perceptions including community engagement; 2) Aligned Values, reflecting non-traditional perceptions of shared values between the U.S. and South Korea; 3) Protection, reflecting the traditional perception of the USFK as a protective security force; and 4) Number of Troops, reflecting traditional perceptions of the scale and presence of the USFK. Our survey also included a measure of respondents’ perception of military conflict in the peninsula, which allows us to explore whether USFK messaging can influence the level of imminent conflict perception among South Koreans. This additional analysis together helps us speculate about the mechanisms through which our information treatments shape the USFK/U.S. favorability among Koreans.

Table 4 presents the regression results using responses to the four statements and conflict perception as dependent variables. Exposure to traditional messaging correlates positively with all four statements (Models 9-12), highlighting potentially diverse mechanisms that such messaging can improve the image of the USFK and the U.S. among South Koreans. Conversely, exposure to non-traditional messaging is only significant for Protection (Model 11). The lack of statistical significance in Civil Society (Model 9) and Aligned Values (Model 10) for the non-traditional group compared to the traditional group is notable, as we had anticipated stronger correlations for the non-traditional group with these variables. Particularly striking is the larger coefficient for Protection in the non-traditional group compared to all other significant coefficients. This suggests that non-traditional messaging may have fostered perceptions of the USFK as a more welcomed protector of South Korea.

Lastly, Model 13 indicates that exposure to non-traditional messaging is significantly negatively associated with fears of conflict on the Korean Peninsula, suggesting that such messaging, particularly containing innocuous images of cultural and sport exchanges, might mitigate perceived conflict risks. In contrast, traditional messaging centered on security and featuring images of military equipment does not appear to influence conflict perception.¹¹ Models 11 and 13 together suggest that if the risk of military conflict on the Korean Peninsula is to be downplayed while maintaining the image of the USFK as a protector of the South Korean public, then employing a non-traditional messaging strategy could be effective.

Open-ended opinion analysis

The final survey question was optional and open-ended, allowing respondents to share additional thoughts. Thirty-eight relevant responses were received. While the limited sample size prevents us from drawing conclusions about the treatment effect, these responses offer valuable insights that could inform future research in this area.

We divided the responses into proponents and critics. The proponent comments center on two themes: personal feelings of (in)security and advocacy for a strong ROK-U.S. alliance. The former refers to personal feelings of safety regarding the presence of the USFK and the hope of no war occurring on the Korean peninsula. One respondent succinctly put, “I feel secure because there are U.S. troops stationed in South Korea.”¹² As for the advocacy for a stronger ROK-U.S. alliance, respondents commented on maintaining, if not improving, a firm military alliance with the U.S. for the sake of national security in the form of continuing the stationing of U.S. troops on South Korean soil. As one respondent wrote, “U.S. troops stationed in South Korea must be stationed to ensure peace and deter war on the Korean Peninsula.” Several used indefinite language, such as “eternal” and “forever” to describe the ROK-U.S. alliance, signifying their perceived necessity of the USFK for the unforeseeable future.

Critics wrote mainly on two themes: security exploitation and military self-reliance. Comments on security exploitation concern the U.S. using South Korea for its own security interests in the region. As two respondents put, “The United States uses South Korea” and “The United States should not drag South Korea into the risk of a new Cold War.” From this, it can be seen that some South Korean citizens feel as if their country is being dragged into conflict that would have otherwise not occurred if the South Korean military was fully self-reliant, or if there was greater equality in the ROK-U.S. alliance in partnership, with this also being mentioned as a concern in some literature (Kim and Boas 2019; Yeo 2011). As military self-reliance, responses emphasized the preference for an autonomous South Korean military and that the phased withdrawal of the USFK is necessary to minimize reliance on the U.S. Two responses read, “It is good to have U.S. troops stationed in South Korea together for security purposes, but we should not rely on them too much,” and “We must be able to protect our country on our own.” Phrasing from these responses seems to signify that the current stationing of the USFK is justified, but over time, a fully self-reliant South Korean military is ideal. This could align with the insignificant findings for Number of Troops in Table 4 and is also consistent with rhetoric from South Korean political elites over the years (Rich et al. 2022; Yeo 2011).

A breakdown of responses between the three experimental groups is visually depicted in Figure 2. It shows that more respondents in the traditional group were willing to leave a response, and more respondents in both treatment groups left comments in support of the USFK compared to the control group.

To gauge the effectiveness of the U.S. military’s digital public diplomacy in shaping public opinion in an allied host country, we conducted a survey experiment exposing South Korean citizens to two variations of messaging: one highlighting traditional military functions and the other focusing on non-traditional functions. We found that both types of military messaging significantly and positively shape perceptions of the U.S. as well as the USFK in South Korea. Differences in effect sizes between these messaging types were not substantial. An unexpected additional finding was that the non-traditional treatment is significantly and negatively associated with fears of conflict on the Korean Peninsula in the near future.

Overall, this research provides compelling evidence that U.S. military messaging has a positive impact on changing perceptions of both the U.S. and the USFK in South Korea. This study thus makes a valuable contribution to the field of military public diplomacy, highlighting the importance of military messaging in shaping public perceptions. As noted, the U.S. military has not been proactive in digital public diplomacy. Presently, @USForcesKorea’s audience demographics are composed of largely English-speaking people, and the account primarily tweets in English. However, as mentioned previously, tweets translated into Korean have more interactions, demonstrating that there is a South Korean audience for the USFK’s public diplomacy. This study’s findings can serve as an effective justification for @USForcesKorea to introduce more Korean-language content into their feed and foster a more Korean-friendly online space to better achieve their public diplomacy goals.

This study has limitations. Due to the volatile nature of opinion surveys, replicating our survey could be challenging. As noted, perceptions of the USFK are influenced by various contextual factors, such as the level of security threat, diplomatic relations with the U.S., and events shaping local NIMBY sentiments concerning U.S. bases. Our survey was conducted in October 2023, when the allied relationship was strong¹³ and there were no salient triggers for anti-American sentiment was present. Considering existing studies’ findings that diplomatic tensions with target countries render public diplomacy efforts ineffective or even counter-effective (Mattingly and Sundquist 2021; Goldsmith et al. 2014), our findings might have differed during low points in the alliance relationship, such as in Fall 2017, immediately after U.S. President Trump expressed his distaste for allies.

Additionally, this research focused specifically on South Korea, and the generalizability of the findings to other U.S. allies or military alliance contexts is uncertain. The impact of non-traditional messaging observed in our study might be less replicable in host countries where U.S. troops’ integration with local society is already strong and citizen attitudes are generally more favorable, as the room for improvement might be smaller. The generalizability of our findings may also be affected by the hosting country’s geopolitical situation. In regions experiencing heightened security threats or political instability, the traditional role of the military as a defense and security apparatus may become more prominent, potentially overshadowing the non-traditional functions, making it more challenging to replicate the positive effects of non-traditional messaging observed in our study.

Nevertheless, we believe this research opens several exciting avenues for future inquiry. First, replicating the study at other times or in other countries with a U.S. troop presence, including cases in similar situations like Japan, where NIMBY sentiments are similarly strong, or in Europe, where regional security situations and sentiments differ greatly, could provide valuable comparative insights and help more comprehensively understand the underlying mechanisms driving public opinion across different settings. Moreover, designing rigorous empirical research exploring specific respondent characteristics not measured in our survey, such as whether respondents served in the military as part of the Korean Augmentation to the United States Army (KATUSA) program or if they know a member of the USFK personally, could help identify and target demographics most susceptible to different public diplomacy messaging strategies. Another interesting avenue for future research would be to explore the effect of different themes of non-traditional messages. While our treatment focused on cultural exchanges, future research could examine whether highlighting the positive economic contributions of U.S. bases could improve perceptions of the U.S. military and the U.S. more broadly.

Lastly, the open-ended responses raise intriguing questions that could not be fully addressed in this paper. While research on the opinions of the future of the USFK has been conducted (Snyder et al. 2019), further investigation into how the U.S. military’s public diplomacy might influence South Korean citizens’ views on the potential withdrawal of the USFK and the development of a self-reliant South Korean military could provide valuable insights for scholars and practitioners deliberating the future security strategies of both the U.S. and South Korea.