Introduction

In modern international relations, nations increasingly employ public diplomacy to build soft power, enhancing their global influence through attraction and persuasion rather than coercion (Nye 2004; Cull 2009; Kelechi 2024). Within this approach, cultural symbols — whether artistic expressions, historical narratives, or iconic representations of national identity — serve as powerful conduits of meaning, conveying cultural values and shaping international perceptions (Hall et al. 2024; Anholt 2007; Jowett and O’Donnell 2018; Yang et al. 2020; Seyfi et al. 2020; Radomskaya and Pearce 2021). Within this framework, animal diplomacy has emerged as a distinct strategy, utilizing charismatic species to advance diplomatic objectives (Zreik 2022).

China’s “panda diplomacy” represents one of the most successful and enduring examples of animal diplomacy (Buckingham et al. 2013; Songster 2018; Longhi and Morelli 2025). The practice has evolved from simple goodwill gifts to sophisticated cooperative breeding and research programs, with pandas loaned to foreign nations. These initiatives serve multiple objectives: fostering international goodwill, strengthening bilateral relations, promoting cultural exchange, and projecting China’s soft power globally. Giant pandas’ effectiveness as diplomatic agents stems from their universal appeal. Revered as national treasures in China and globally recognized as conservation icons, pandas consistently generate substantial public interest and media coverage in host countries (Songster 2018; Huang and Wang 2020). This widespread appeal transcends cultural and political boundaries, potentially influencing international tourism. The hypothesis suggests that positive media exposure and public engagement with pandas in host countries can stimulate interest in China as a destination (Huang and Wang 2020).

Despite widespread recognition of cultural symbols’ influence on international tourism, quantitative research measuring these effects remains scarce. The existing literature primarily consists of qualitative discussions on how cultural symbols shape national image and foster cultural exchange, often lacking systematic quantitative analysis. For instance, Radomskaya and Pearce (2021) introduced the concept of destination mascots, demonstrating their role in enhancing tourism through place identity markers and cultural constructs. Previous quantitative studies examining cultural symbols’ influence have faced several significant limitations. These studies are often restricted by limited data availability, small sample sizes, and a focus on single countries, which constrains their generalizability and broader applicability for policymakers. This limitation is exemplified in Seyfi et al.‘s (2020) study, which developed a theoretical model of cultural tourism experiences based on interviews with a small sample of tourists in Paris. Two primary challenges have hindered comprehensive research in this field. The first challenge involves the difficulty in obtaining comprehensive datasets suitable for detailed econometric analysis. The second challenge stems from the localized nature of many cultural symbols, which complicates global recognition and tracking of their impacts. The giant panda presents a unique opportunity for studying cultural symbols’ global influence. This exceptional case stems from the panda’s widespread international appeal, strong association with China, and well-documented presence in host countries. The significant public exposure and trackable impact across multiple nations makes the giant panda an ideal subject for examining the broader impact of cultural symbols on international tourism.

This research addresses the empirical gap in cultural symbol impact studies through several methodological approaches. The study employs a comprehensive cross-national dataset, applies rigorous econometric methods, and conducts quantitative analysis of the giant panda’s effect on tourism flows. The study evaluates the influence of China’s Panda Cooperative Breeding Program (CPCBP) on international tourist arrivals to China. This evaluation examines the direct impact on tourism flows, the contribution to national image enhancement, and the economic and cultural benefits of the initiative. The study addresses two primary research questions. First, it examines how CPCBP affects international tourism to China. Second, it investigates which specific program aspects most effectively drive tourism increases, including the location effects of panda exhibitions and the timing impacts of significant events such as panda births.

This study advances both academic understanding and practical policy by quantifying the causal relationship between “panda diplomacy” and China’s inbound tourism. This research makes three key contributions. First, it moves beyond qualitative assessments of soft power by developing a robust model to measure the economic returns of public diplomacy initiatives. Second, recognizing panda loans as strategic foreign policy instruments, it evaluates their concrete tourism dividend. Third, it examines how China leverages the giant panda—a powerful cultural symbol—to transform symbolic capital into measurable economic benefits through tourism. The analytical framework integrates multiple theoretical perspectives: national image theory (Anholt 2007; Wang 2006), cultural symbolism (Hall et al. 2024; Geertz 2017), animal diplomacy (Leira and Neumann 2017), and models of international tourism demand (Song and Li 2008; Crouch 1994). This approach distinguishes the present research from studies focused on localized impacts, such as zoo attendance or reverse tourism flows, including Chinese outbound tourism to panda-recipient countries (Okafor et al. 2021). Instead, it examines the macroeconomic impact on China’s inbound tourism, investigating how this form of public diplomacy influences aggregate international tourist arrivals from panda-hosting nations.

The research methodology employs three complementary approaches to ensure robust and reliable results. The primary analytical tool is a Staggered Difference-in-Differences (DID) model, which establishes causal relationships across time and contexts (Callaway and Sant’ Anna 2021; Athey and Imbens 2022; Roth et al. 2023). To address potential endogeneity between the program and tourism flows, we implement an instrumental variable two-stage least squares (IV-2SLS) approach. Additionally, this study employs a Propensity Score Matching Difference-in-Differences (PSM-DID) method to enhance validity by matching countries based on observable characteristics. This comprehensive methodology, applied to data from 158 countries between 2004 and 2019, provides robust insights into the relationship between cultural diplomacy and international tourism flows.

The remainder of this paper is organized as follows. Section “Literature Review” reviews the literature on public diplomacy, animal diplomacy, and tourism impacts, with particular attention to China’s panda diplomacy program. Section “Identification strategy, variables and data” details the methodological approach, presenting the identification strategy and describing the comprehensive dataset covering 158 countries from 2004 to 2019. Section “Empirical Analysis” presents the empirical findings and discusses their implications for understanding the relationship between cultural diplomacy and tourism flows. Finally, Section “Discussion and Conclusion“ summarizes the key findings, discusses policy implications, and suggests directions for future research.

Literature Review

This section synthesizes multiple streams of literature to establish the theoretical foundation for examining how panda diplomacy influences China’s inbound tourism. The review integrates five key research areas: public diplomacy theory, national image formation, cultural symbolism in international relations, animal diplomacy practices, and determinants of international tourism flows. The integration of these interconnected bodies of literature provides a comprehensive framework for understanding how China’s panda diplomacy program functions as a public diplomacy tool to influence international tourism patterns.

Public diplomacy, national image, and cultural symbols

Public diplomacy has become increasingly vital in international relations, distinguished from traditional diplomacy by its direct engagement with foreign publics through cultural exchanges, educational programs, and information sharing (Cull 2009; Nye 2004; Gilboa 2008; Hartig 2016). At its core, public diplomacy aims to shape and project a positive national image to advance long-term national interests (Anholt 2007). National image—the collective beliefs and impressions people hold about a country—comprises three key dimensions (Kunczik 2016; Anholt 2007). The cognitive dimension encompasses knowledge about a country’s attributes, such as economic strength and cultural heritage. The affective dimension involves emotional responses to a nation, while the conative dimension relates to behavioral intentions, such as decisions to visit or invest (Gartner 1994; Yang et al. 2022).

Cultural symbols serve as powerful conduits for national image construction and cross-cultural communication (Whyke et al. 2022). These symbols function as concentrated representations of national identity and values, transcending linguistic and cultural barriers to convey messages intuitively and vividly (Kaneva 2011; Whyke et al. 2022). This aligns with Nye’s (2004) concept of soft power, where cultural attractiveness becomes a tool for influence through attraction rather than coercion.

Animal symbols, particularly flagship species and charismatic megafauna, have proven especially effective in international communication (Lundberg and Arponen 2022). These species—such as the giant panda, tiger, polar bear, or bald eagle—transcend their biological significance to become cultural icons, capable of projecting specific national traits and fostering positive associations globally (Lorimer 2007; Skibins 2015). However, the effectiveness of national symbols varies across cultural contexts. The Chinese dragon illustrates this challenge: while representing wisdom and good fortune in Chinese culture (Dikötter 1997; Yuan 2015), it often evokes negative associations in Western traditions (Rose 2001; Yuan 2015; Zeng 2008). In contrast, the giant panda has emerged as a uniquely successful cultural symbol for China, enjoying consistently positive recognition across cultures. Its image—associated with gentleness and vulnerability—generates widespread affection and warmth, largely free from conflicting cultural interpretations (Buckingham et al. 2013; Songster 2018; Yang and Lin 2022).

Animal diplomacy and the unique symbolism of the giant panda

Animal diplomacy encompasses the strategic use of animals in international relations through gifts, loans, or joint conservation efforts. This diplomatic practice aims to foster goodwill, strengthen bilateral relationships, and enhance national image by leveraging animals’ intrinsic appeal and symbolic power (Ritvo 1989; Cushing and Markwell 2009; Leira and Neumann 2017). While rooted in historical traditions of rulers exchanging exotic animals as symbols of power and alliance (Ritvo 1989), contemporary animal diplomacy has evolved into a sophisticated tool of public diplomacy. Several nations have employed animal diplomacy through their iconic species. The United States projects its values through the bald eagle’s symbolism, Russia’s international identity is closely tied to the bear (Riabov and de Lazari 2009; de Lazari et al. 2019), and Australia occasionally uses koalas and kangaroos in diplomatic gestures (Cushing and Markwell 2009; Caldwell and Henry 2020; Markwell 2021). These initiatives aim to generate positive media coverage, foster cross-cultural connections, and enhance soft power through natural heritage.

Among these practices, China’s panda diplomacy stands distinct in its scope, longevity, and cultural impact (Songster 2018). The giant panda’s effectiveness as a diplomatic tool stems from unique biological and cultural attributes. Endemic to China and revered as a national treasure, the panda’s appealing features—including its distinctive appearance and peaceful nature—resonate across cultures (Darmawan and Putri 2021; Cho et al. 2022; Lehmann et al. 2013; Sanefuji et al. 2007). Its black-and-white coloring symbolizes harmony, while its bamboo diet aligns with China’s diplomatic philosophy of non-aggression (Songster 2018). The evolution of China’s panda diplomacy reflects its changing diplomatic strategies. Beginning with China's 1941 gift to the United States, the practice was formalized after 1949 under Chairman Mao and Premier Zhou (Hartig, 2013; Songster, 2018). A significant expansion occurred around 1972, coinciding with China’s diplomatic normalization with the U.S. and Japan (Chinoy 1975; Schaller 1994). The program transitioned from outright gifts of 23 pandas to nine countries between 1957–1982 to a sophisticated cooperative breeding model in 2007, emphasizing research collaboration and conservation (Hartig 2013; Darmawan and Putri 2021; Okafor et al. 2021). The panda effect on local tourism in host countries is well-documented (Huang and Wang 2020). Edinburgh Zoo saw visitor numbers increase from 536,000 to 800,000 following the pandas’ arrival in 2011 (BBC News 2012), while Washington’s National Zoo has maintained strong visitor interest since 1972 (Smithsonian’s National Zoo and Conservation Biology Institute, n.d.). While studies like Okafor et al. (2021) have examined Chinese outbound tourism to panda-hosting countries, the reverse effect—how panda diplomacy influences tourism to China from host countries—remains empirically unexplored.

Determinants of international tourism demand

The demand for international tourism is a complex phenomenon shaped by a multitude of interacting factors. The literature broadly categorizes these into economic, political, geographical, and cultural determinants, which collectively influence tourist decisions and shape global tourism flows (Gidebo 2021).

Tourism, as discretionary consumption, demonstrates high demand elasticity (Crouch 1994; Song and Li 2008; Peng et al. 2014). Disposable income in tourist-originating countries significantly influences international travel capacity (Oh 2005; Lanza et al. 2003). Relative prices of tourism services and exchange rates between origin and destination countries also play crucial roles—a stronger home currency typically stimulates outbound tourism by making destinations relatively cheaper (Lim 1997; Song and Li 2008). Studies have identified a bidirectional relationship between tourism and economic growth, with each reinforcing the other (Dritsakis 2004; Durbarry 2004).

Political stability and diplomatic relations fundamentally affect tourism flows (Liu and Pratt 2017; Pradeep 2020; Wintersteiner and Wohlmuther 2014). Boyd’s (2000) study of Northern Ireland demonstrated how reduced political tension increased cross-border tourism. Safety perceptions significantly influence destination choice, with political instability, crime, or terrorism threats deterring visitors (Sönmez and Graefe 1998). Tourism infrastructure—including transportation networks, accommodation, and communication systems—directly impacts destination attractiveness and visitor experience (Prideaux 2000; Mandić et al. 2018).

The gravity model of tourism, adapted from international trade theory, explains how tourism flows relate to economic size and geographical distance (Tinbergen 1962; Anderson and Van Wincoop 2003). Tourism volume typically decreases with distance due to increased travel costs and time (Mckercher, Lew (2003)). This model effectively predicts tourism patterns and explains why neighboring countries often generate significant visitor numbers (Gil-Pareja et al. 2007; Rossello Nadal and Santana Gallego 2022). Cultural distance, measured through Hofstede’s (1984) dimensions and Kogut and Singh’s (1988) composite index, presents both opportunities and challenges. While cultural differences can attract tourists seeking novelty and authenticity (Ng et al. 2007; Liu et al. 2018), they may also deter visitors by increasing perceived risks and psychological stress (Wąsowicz-Zaborek 2025).

Finally, cultural assets significantly influence tourism demand (Noonan 2022). The UNWTO recognizes cultural tourism as driven by visitors seeking cultural experiences. Research confirms that cultural facilities—museums, exhibitions, archaeological sites—enhance tourism performance (Stylianou-Lambert 2011; Di Lascio et al. 2011; Guccio et al. 2017). National treasures, representing significant cultural or historical heritage, form the foundation of heritage tourism, converting cultural capital into tourist flows (Handler 1988; Poria et al. 2003; Timothy and Boyd 2003).

Literature synthesis and research gaps

The review of the literature reveals three key insights. First, cultural symbols function as effective public diplomacy tools for shaping national image, though their impact varies significantly with cross-cultural interpretation, as demonstrated by contrasting global responses to the dragon versus the panda (Nye 2004; Anholt 2007; Rose 2001; Songster 2018). Second, animal diplomacy, particularly China’s panda program, represents a distinctive and strategic approach to fostering international goodwill. Third, international tourism demand responds to multiple determinants, including economic conditions, infrastructure quality, and cultural-geographical variables (Crouch 1994; Ng et al. 2007). The existing research, however, exhibits two significant limitations. Current studies are predominantly historical or descriptive in nature, lacking rigorous quantitative analysis of panda diplomacy’s impact (Songster 2018). Additionally, while Okafor et al. (2021) established a quantitative link between panda loans and increased Chinese outbound tourism to host nations, the reverse relationship—how panda diplomacy affects tourism to China from host countries—remains empirically unexplored.

This study addresses these research gaps through three primary contributions. First, the analysis provides a comprehensive quantitative assessment of panda diplomacy’s impact on China’s inbound tourism. Second, the research integrates public diplomacy theory with established tourism demand models to evaluate how symbolic diplomatic gestures translate into measurable economic outcomes. Third, the study employs a Difference-in-Differences (DID) methodology to establish causality by comparing tourism patterns between panda-receiving countries and similar non-recipient nations before and after panda loans.

Identification strategy, variables and data

Identification strategy

The CPCBP exhibits two key sources of variation: temporal differences in panda loans and births, and cross-sectional differences between host and non-host nations. This variation pattern enables the application of a Staggered Difference-in-Differences (DID) methodology to evaluate the program’s impact on China’s inbound tourism (Callaway and Sant’Anna 2021; Athey and Imbens 2022; Roth et al. 2023). The program’s implementation between 2004 and 2019 created a series of natural experiments through its systematic expansion. During this period, China loaned 29 pandas to 14 countries, while 46 panda cubs were born and raised in foreign locations. In total, 96 pandas resided across 16 countries throughout the study period, creating a substantial network of international panda presence. This implementation structure creates an ideal analytical framework for examining the program’s impact. The 14 countries that received panda loans form the treatment group, while the remaining countries constitute the control group. The staggered timing of loans to different countries generates distinct policy shocks, providing multiple points of comparison for analyzing the program’s effects on tourism patterns.

$$touris{t}_{it}={\beta }_{0}+{\beta }_{1}pandaprogra{m}_{it}+\sum _{j}{\beta }_{j}control+{\gamma }_{t}+{\mu }_{i}+{\varepsilon }_{it}$$

In this model, i and t denote country and year, respectively. The dependent variable touristit represents the number of tourists from country i visiting China in year t. The key independent variable pandaprogramit indicates whether China loaned pandas to country i in year t. Control variables are denoted by Xit; γt represents year fixed effects, and μi represents country fixed effects, which control for unobserved factors varying across year and countries. The coefficient β1 captures the net effect of pandas on inbound tourism to China.

Variables

The dependent variable in this study is touristit, indicating the number of tourists from country i visiting China in year t. The key independent variable is pandaprogramit, which is set to 1 from year t onwards if pandas were loaned to country i in year t, and 0 otherwise.

Considering various factors influencing international tourism, we include several control variables based on relevant literature and practical considerations:

Population size (pop) as a proxy for market size.

Log of per capita national income (lngni) as a proxy for income level.

Ratio of purchasing power parity conversion factor to market exchange rate (relativeprice) as a proxy for relative prices.

Ratio of merchandise trade to gross national income (traderatio) as a proxy for the degree of openness.

Information flow indicator from the KOF Globalization Index (infrastructure), including television, internet, and newspaper circulation per GDP.

Financial development indicator from the KOF Globalization Index (finance).

Political stability and absence of violence index (polistable).

Data sources and descriptive statistics

This study draws upon annual panel data encompassing 158 countries from 2004 to 2019. This specific timeframe was selected based on three critical considerations: First, data quality and availability shaped our temporal boundaries. The year 2004 marks the beginning of comprehensive and consistent tourism statistics across our panel of 158 countries. The analysis concludes in 2019 to avoid the unprecedented disruption to global tourism caused by the COVID-19 pandemic. Second, this period captures a significant era in China’s tourism development. Following China’s accession to the World Trade Organization (WTO) in 2001, international tourism experienced accelerated growth. Beginning the analysis in 2004 allows for a focus on this period of sustained expansion. Third, the 2004 start date strategically avoids the temporary but severe disruption caused by the 2003 SARS outbreak, enabling the model to capture more typical patterns in international tourism flows. Data for the dependent variable, touristit, are obtained from the World Tourism Organization and the China Tourism Statistical Yearbook. The primary independent variable data are compiled from Chinese official news reports, the Chengdu Research Base of Giant Panda Breeding, and the Panda Profiles website. Additionally, supplementary data are sourced from the World Bank, the International Monetary Fund, and the French Centre for Research and Prospective Studies.

Detailed variable information and descriptive statistics are presented in Table 1.

Table 1 Variable Information and Descriptive Statistics.
Full size table

Empirical Analysis

Correlation analysis

To establish initial evidence of the CPCBP’s impact on inbound tourism, the analysis employs the Synthetic Control Method. This analytical approach, developed for policy evaluation, creates a data-driven comparison group by constructing a weighted combination of untreated units (Abadie et al. 2010; Abadie 2021). The method operates by creating a synthetic control unit that closely matches the treated unit’s pre-intervention tourism patterns through randomized simulation experiments. This synthetic counterpart enables causal inference by providing a baseline for comparing post-intervention outcomes (Arkhangelsky et al. (2021)). While the Synthetic Control Method offers valuable insights, its limitation lies in analyzing only one treatment unit at a time. Therefore, this analysis uses it as a preliminary analytical tool to complement our main empirical strategy—the Difference-in-Differences approach (Roth et al. 2023).

The synthetic control analysis employs several predictor variables to construct appropriate comparison groups. The selected predictor variables include per capita national income, information infrastructure, economic openness, geographical distance, cultural distance, institutional distance, and inbound tourism numbers from the two years prior to program initiation. These predictors enable the assignment of appropriate weights to non-panda-hosting countries, thereby creating synthetic control groups that serve as benchmarks for comparing actual inbound tourism numbers. The selection of cases for Synthetic Control Method (SCM) analysis adheres to strict methodological criteria. As Abadie (2021) emphasizes, the credibility of SCM findings depends on having both a sufficient pre-treatment period to establish a well-fitted synthetic control and an adequate post-treatment period to assess intervention impacts. Given our 2004–2019 sample period, the analysis excludes countries that initiated panda cooperation either too early (near 2004) or too late (near 2019), as these cases would lack either sufficient baseline data or follow-up periods for reliable analysis. Following these methodological requirements, four representative countries were selected whose panda programs began during the middle of the sample period: Australia (2009), the United Kingdom (2011), Spain (2013), and Canada (2014). These cases provide balanced timeframes that allow for both the construction of credible counterfactuals and the evaluation of program effects with statistical rigor.

Figures 1 through 4 illustrate the synthetic control analysis outcomes for these representative countries. The results show that before the program’s initiation, trends in actual inbound tourism (solid lines) and synthetic inbound tourism (dashed lines) generally align, indicating that the synthetic control groups serve as appropriate counterfactuals for the treatment groups. After the program begins, the trends diverge, with actual inbound tourism numbers significantly surpassing those of the synthetic control. This divergence suggests that the program might positively influence inbound tourism to China, although more rigorous econometric modeling is needed to confirm a causal relationship.

Fig. 1
Fig. 1
Full size image

Synthetic Control Results for Australia (Start Year: 2009).

Fig. 2
Fig. 2
Full size image

Synthetic Control Results for the U.K. (Start Year: 2011).

Fig. 3
Fig. 3
Full size image

Synthetic Control Results for Spain (Start Year: 2013).

Fig. 4
Fig. 4
Full size image

Synthetic Control Results for Canada (Start Year: 2014).

Baseline regression results analysis

Overall effect

The impact evaluation of the panda breeding program on Chinese tourism employed a Staggered Difference-in-Differences (DID) model. The baseline regression results, presented in Column (1) of Table 2, demonstrate a statistically significant positive coefficient for the variable pandaprogram at the 1% level, after controlling for country and year fixed effects. This indicates that panda loans significantly increase tourist flows to China.

Table 2 Baseline Regressions and Lagged Effects.
Full size table

Lagged effect

Considering that pandas require time to acclimate to their new environment and accounting for a potential delay in public engagement and outbound tourism response, it is hypothesized that the program’s impact will manifest with some lag. To examine this hypothesis, lagged variables representing the panda breeding program were introduced into the model. The regression results, displayed in Columns (2) to (5) of Table 2, reveal that the coefficients for lagged variables from one to four years are significantly positive, with the most substantial effect occurring in the second year after the program’s initiation. This impact gradually decreases in both magnitude and significance over time. This finding corroborates the hypothesis that the panda breeding program’s positive influence on tourism to China exhibits a lagged effect, which diminishes over time.

Effect of Panda cub births

Panda cub births generate substantial public interest, often exceeding the attention received by adult pandas. This phenomenon is exemplified by Xiang Xiang, born at Ueno Zoo in Japan, who achieved iconic status. Her popularity was so significant that her return to China faced five postponements due to public demand and COVID-19 considerations, with thousands of visitors attending her final day at the zoo.

To test the effect of panda cub births, a dummy variable babybornit was introduced, which equals 1 if a panda cub is born in country i in year t, and remains 1 for all subsequent years until 2019. For countries with multiple births, the first cub’s birth year served as the event marker. The regression results in Column (6) of Table 2 show a significant positive coefficient at the 5% level for the variable babyborn confirming that cub births contribute substantially to increased Chinese inbound tourism.

Extended analysis: impact of panda cub returns and deaths

Beyond examining panda births, the analysis investigates how two additional events—the return of cubs to China and the death of pandas abroad—influence tourism patterns. This investigation is motivated by the significant public attention generated by high-profile panda returns, such as Xiang Xiang from Japan and Fubao from South Korea, which potentially redirect public interest toward China. To test these hypotheses, the analysis introduces two new variables to the model. The first variable, babybackit, is a dummy variable set to 1 for country i in year t when the first foreign-born cub returns to China, and remains 1 for all subsequent years. It focuses on the first returning cub to maintain methodological consistency with the earlier DID analysis of cub births. The second variable, pandadeathit, is a dummy variable coded as 1 for country i in year t if a loaned panda dies, and remains 1 thereafter.

The regression results, presented in Columns (7) and (8) of Table 2, reveal distinct effects for these events. The coefficient for babyback shows positive and statistically significant impact on inbound tourism at the 10% level (Column 7). This finding suggests that cub returns generate tangible increases in tourism from host countries, likely due to sustained media coverage and emotional connections formed during the cub’s stay abroad (Huang and Wang 2020). These return events appear to reinforce China’s cultural connection with pandas, potentially motivating international visitors to follow the cubs to their new home. While several high-profile returns, including Xiang Xiang from Japan and Fu Bao from South Korea, fall outside our study period (2004–2019), they present valuable opportunities for future research examining the panda effect in the post-pandemic tourism landscape.

In contrast, the coefficient for pandadeath is negative but statistically insignificant (Column 8). This result likely reflects data limitations rather than the absence of an effect. The dataset captures only two cases meeting DID criteria: the death of Long Hui in Austria (2016) and the deaths of Ling Ling and Mei Mei in Japan (2008). Other panda deaths either occurred outside our observation window or involved pandas not covered by the DID framework, thereby limiting the statistical power to detect potential effects.

Endogeneity

Endogeneity poses potential challenges in estimating the impact of the panda breeding program on international tourism flows. While the baseline regressions incorporate country and year fixed effects to address omitted variable bias and country-specific selection effects, simultaneity bias remains a concern. Specifically, the relationship between program implementation and tourism flows may exhibit bidirectional causality, as countries with growing tourism potential might be more likely to receive pandas, while the presence of pandas could simultaneously influence tourism patterns.

To address the issue of simultaneity, the instrumental variable two-stage least squares (IV-2SLS) method was applied. The variable country82, indicating whether a country received pandas as gifts from China before 1982, served as the instrumental variable. Prior to 1982, China gifted pandas at a high cost. To improve panda conservation, China stopped gifting pandas in 1982 and transitioned to cooperative research endeavors in the 21st century, choosing partner countries based on their expertise and capacity to care for pandas. This historical context establishes a significant correlation between countries that received pandas before 1982 and those that entered into panda cooperation with China in the 21st century. In other words, countries that originally received pandas are expected to also be involved in recent cooperative efforts, establishing the variable’s relevance. Regarding exclusivity, since China initiated its opening-up policy in 1978 and joined the World Trade Organization in 2001, it is plausible to assume that post-2004 international tourism activities are not directly tied to countries that received pandas almost four decades ago. Therefore, given its relevance and exclusivity, this instrument is deemed appropriate.

As shown in Table 3, in the first stage of the 2SLS regression, the instrumental variable country82 positively influences the independent variable pandaprogram at the 1% significance level. Both the Anderson LM statistic and the Cragg-Donald Wald F statistic for weak instruments surpass critical thresholds, confirming the instrument’s validity. In the second stage, after accounting for potential endogeneity, the regression coefficient for pandaprogram remains significant at the 1% level, indicating that the program has a robust positive effect on inbound tourism to China. These results reinforce the assertion that the panda breeding program significantly enhances tourism to China, illustrating the effectiveness of utilizing cultural symbols for international tourism promotion.

Table 3 Instrumental Variable 2SLS Regression Results.
Full size table

Robustness checks

Parallel trends test and dynamic effects

The validity of the Difference-in-Differences (DID) methodology rests fundamentally on the parallel trends assumption (Callaway and Sant’Anna 2021; Athey and Imbens 2022; Roth et al. 2023). Under this assumption, in the absence of panda diplomacy, tourism trends to China should follow similar patterns between panda-hosting countries (treatment group) and non-hosting countries (control group). The verification of this assumption is crucial for establishing causality, as pre-existing differences in tourism trends between treatment and control groups would compromise the credibility of the estimated treatment effects.

To test this assumption, the methodology outlined by Ferrara et al. (2012) was employed. Dummy variables were introduced for each of the five years before and after the program, serving as explanatory variables in the model. For countries without pandas during the sample period (e.g., the USA and Austria), the birth of the first panda cub was designated as the shock time. The regression coefficients and their 95% confidence intervals were then plotted.

Figure 5 illustrates that the coefficients for the years preceding the panda program are not statistically significant at the 95% confidence interval, indicating that there were no systematic differences in tourism trends between countries that did and did not host pandas. Following the program’s initiation, the coefficients for the second year are significantly positive at the 5% level, while those for subsequent years fail to show significance. This finding corroborates earlier results, suggesting that the impact of the pandas is delayed.

Fig. 5
Fig. 5
Full size image

Parallel Trends Test.

Placebo test

The stringent criteria required for hosting pandas, such as advanced animal research capabilities, high economic development, and favorable climate conditions, mean that host countries are typically developed nations or those with strong trade ties to China. This introduces concerns about potential selection bias and policy endogeneity, which could undermine the reliability of the DID regression results (Callaway and Sant’Anna 2021; Athey and Imbens 2022; Roth et al. 2023). To address these concerns, a placebo test was conducted, following the methodology of Ferrara et al. (2012). In this test, 14 countries were randomly selected from a pool of 158 to form a “pseudo-treatment group”, while the others were designated as the “pseudo-control group”. The same regression analysis was applied to these groups. If the coefficients from the placebo analyses are not significantly different from zero, it would suggest that there is no selection bias in the original group assignments.

Figure 6 illustrates the probability density distribution of coefficients from 500 random iterations, with the vertical line marking the actual estimated coefficient. The distribution centers around zero, while the actual coefficient significantly deviates from this distribution. This result suggests that the initial findings are unlikely to be biased by selection issues.

Fig. 6
Fig. 6
Full size image

Placebo Test.

Propensity Score Matching Difference-in-Differences (PSM-DID)

To further mitigate potential biases arising from data and model selection, the Propensity Score Matching Difference-in-Differences (PSM-DID) approach was employed (Blundell and Costa Dias 2000; Heyman et al. 2007). This method enhances the validity of comparisons by integrating Propensity Score Matching (PSM) with the DID approach, thereby matching countries based on multiple observable characteristics to create a control group similar to the treatment group (Heyman et al. 2007). Based on relevant literature, countries were matched using six observable variables: per capita national income, information infrastructure, openness, geographical distance, cultural distance, and institutional distance. A Logit model was used to estimate the likelihood of a country successfully borrowing pandas from China. Nearest-neighbor matching with a 1:4 ratio was then applied to ensure robust matches before conducting the regression analysis again. The results, shown in Table 4, confirm that the panda breeding program significantly promotes inbound tourism to China, revealing a short-term lag and a diminishing effect over time, consistent with previous findings.

Table 4 PSM-DID Regression Results and Other Robustness Tests.
Full size table

Additional robustness tests

To ensure the robustness of the results, the primary explanatory variable pandaprogram was replaced with the total number of pandas residing in a country (pandanumber), capturing the cumulative effect of panda loans, cub births, deaths, and returns to China. The Ordinary Least Squares (OLS) regression results (see Table 4) show that an increase in the number of pandas significantly enhances tourism to China, supporting earlier analyses. Furthermore, to prevent potential interference from countries that hosted pandas prior to 2004 (Japan, the USA, and Austria), these countries were excluded from the sample, and the regressions were re-run. The results remained consistent with previous findings.

Heterogeneity analysis: host country characteristics and the panda effect

While the main findings confirm panda diplomacy’s positive influence on tourism, the effectiveness of public diplomacy initiatives often varies across national contexts. This section examines how specific characteristics of host countries moderate the “panda effect”, focusing on three key dimensions: economic and cultural openness, governance capability, and cultural distance from China. The key moderating variables are operationalized using established international indices to ensure measurement reliability and comparability across countries. The first moderating variable, openness, is measured using the KOF Globalization Index, which provides a comprehensive assessment of a country’s economic integration, social connectivity, and political engagement with the global community. Within this index, lower scores indicate less openness to international exchange and interaction. The second moderating variable, governance capability, is evaluated using the World Bank’s Worldwide Governance Indicators (WGI). These indicators aggregate multiple assessments of governance quality, including government effectiveness, rule of law, and regulatory quality. Higher scores on the WGI reflect stronger institutional capacity and more effective governance systems. The heterogeneous effects of these host country characteristics on panda diplomacy’s tourism impact are presented in Table 5, revealing the conditions under which panda diplomacy yields the greatest returns for tourism promotion.

Table 5 Subgroup Regression Results.
Full size table

The analysis reveals three key findings about how host country characteristics moderate the panda effect on tourism flows. The panda program demonstrates stronger tourism effects in countries with lower levels of openness, a finding that initially appears counterintuitive. This result can be understood through the lens of communication theory and signal salience in public diplomacy. In highly open societies, citizens face constant exposure to international information, cultural products, and global narratives. Within this saturated environment, even compelling cultural symbols like pandas must compete for limited public attention, creating what Nye (2004) and Simon (1996) term the “paradox of plenty”.

In contrast, countries with more restricted information flows offer a different dynamic. The arrival of giant pandas in these nations becomes a significant national event, as the novelty and scarcity of such a positive, apolitical symbol from a major world power generate sustained media coverage and public discourse. This heightened attention, amplified by a less-cluttered media landscape, more effectively translates into tourism intentions. The finding aligns with research showing that public diplomacy’s impact depends heavily on the target nation’s domestic information environment (Cull 2009). The second key finding reveals stronger panda diplomacy effects in countries with robust governance capabilities. This result highlights how a soft power initiative’s success depends not only on the symbol’s appeal but also on the recipient nation’s state capacity to amplify its impact (Fukuyama 2004).

Countries with strong governance, such as Singapore, Germany, and Canada, possess the institutional infrastructure to maximize panda hosting opportunities. These nations can effectively construct world-class enclosures, launch sophisticated marketing campaigns, develop integrated educational programs, and ensure enhanced visitor experiences. Their well-governed institutions effectively “co-brand” the panda experience, creating synergistic effects that boost tourism. In contrast, countries with weaker governance may struggle with logistical challenges, inadequate promotion funding, or bureaucratic inefficiencies, limiting their ability to translate public interest into tourism growth (Hall et al. 2024).

The analysis of cultural distance reveals significant positive effects across both culturally distant and proximate countries, demonstrating the panda’s universal appeal. While countries with greater cultural distance show slightly larger effects, possibly due to novelty value, both groups exhibit statistically significant positive impacts on tourism flows.

To sum up, these findings demonstrate that panda diplomacy’s effectiveness varies significantly with host country context. The program’s impact is strongest not in the most globally integrated societies, but in environments where its message faces less competition for attention. Additionally, the success of panda diplomacy depends critically on the host nation’s institutional capacity to leverage the initial diplomatic gesture into lasting impact. These insights suggest that cultural diplomacy strategies should be carefully tailored to each target nation’s information environment and institutional capabilities.

Discussion and Conclusion

Discussion

This research examines the impact of CPCBP on international tourism flows to China, utilizing econometric methods to quantify this relationship. By employing a Staggered Difference-in-Differences (DID) model, along with supplementary robustness checks, the study provides empirical evidence on how cultural diplomacy initiatives, particularly involving giant pandas, influence global tourism patterns. The analysis covers data from 2004 to 2019 across 158 countries, allowing for a thorough assessment of the program’s effectiveness. The findings of this study indicate that CPCBP significantly boosts international tourism to China. The key findings are:

Firstly, the program has a statistically significant positive effect on international tourist numbers to China. The staggered DID analysis, combined with robustness checks and heterogeneity analysis, reveals that countries hosting pandas see a notable increase in tourism compared to those that do not. Secondly, the impact of introducing pandas on tourism is delayed. The most substantial increase in tourist numbers occurs in the second year after pandas are introduced, with the effect gradually diminishing over subsequent years. Thirdly, the birth of panda cubs generates a substantial additional boost to tourism, especially in countries where pandas are major attractions. The case of Xiang Xiang in Japan exemplifies this phenomenon, as the birth of a cub significantly elevated public interest and extended the panda program’s positive effects on tourism. These findings remain robust through comprehensive robustness checks.

The heterogeneity analysis reveals the varying impact of the panda program across different countries, underscoring the interplay between cultural diplomacy and national characteristics. The study finds that the program’s effectiveness is markedly increased in countries with lower levels of openness, where international cultural exchanges are rare. In these environments, pandas as cultural symbols capture substantial public interest and extensive media coverage, resulting in a more pronounced increase in tourism. This highlights the importance of novelty and exclusivity in cultural diplomacy, where the rarity of such exchanges amplifies their influence.

Moreover, the analysis underscores the critical role of governance capacity in maximizing the benefits of the panda program. Countries with robust governance structures are better equipped to manage and promote pandas effectively, organizing related events and maintaining public interest over time. These nations can leverage pandas not just as attractions, but as ongoing symbols of cultural exchange, ensuring sustained tourism benefits. The study suggests that the success of diplomatic initiatives is not solely dependent on the intrinsic appeal of the cultural symbol, but also on the host country’s ability to capitalize on it through effective governance and strategic promotion.

In summary, the findings provide robust empirical evidence that CPCBP significantly enhances inbound tourism to China, especially in countries with lower openness and stronger governance capacities. These results underscore the effectiveness of using cultural symbols in international tourism promotion and suggest that tailored strategies that consider host country characteristics can maximize the impact of such programs.

Theoretical contribution and practical implications

This research advances the scholarly understanding of cultural diplomacy and international tourism in three significant ways. First, it moves beyond theoretical discussions of cultural symbols’ value by providing rigorous quantitative evidence of their economic impact. Through the analysis of the giant panda as a case study, the research demonstrates how cultural diplomacy can generate measurable increases in international tourism flows. Second, the study enhances the methodological toolkit for evaluating cultural diplomacy initiatives. The analytical framework combines three sophisticated approaches: staggered difference-in-differences analysis, instrumental variable methods, and propensity score matching. This integrated methodology provides researchers with a more robust approach for measuring the causal effects of cultural diplomacy programs. Third, the research illuminates the mechanisms through which cultural initiatives succeed across different national contexts. By examining how host country characteristics moderate the panda effect, the analysis contributes to a deeper theoretical understanding of cultural diplomacy’s contextual effectiveness.

For policymakers and tourism stakeholders, the findings offer concrete guidance for designing and implementing cultural diplomacy initiatives. The documented success of the CPCBP suggests that strategically deployed cultural symbols can effectively promote tourism while strengthening international relations. The heterogeneity analysis yields particularly valuable insights for practitioners in the field. The finding that program effectiveness varies with host country characteristics challenges the conventional one-size-fits-all approach to cultural diplomacy. The results indicate that initiatives should be tailored to specific national contexts, with particular attention to information environment openness, institutional governance capacity, and cultural distance factors. These findings suggest that practitioners should develop customized strategies that account for each target nation’s unique attributes and capabilities. Such tailored approaches are likely to yield significantly stronger results than standardized strategies that ignore national differences.

Broader relevance and applicability

While the giant panda represents an exceptionally symbolic and emotionally resonant figure in China’s cultural diplomacy, the insights from this study have broader implications for other forms of animal diplomacy and cultural exchange. The identified mechanisms—emotional engagement, media amplification, symbolic storytelling, and the importance of local governance capacity—are not unique to the panda case but are likely applicable to other soft power initiatives that rely on culturally significant or emotionally compelling symbols.

In this regard, the findings suggest three key recommendations for the future design and implementation of public diplomacy strategies. First, nations should actively leverage cultural and public diplomacy to shape national image and promote civilizational dialogue in an increasingly complex global environment. In times of geopolitical uncertainty, initiatives like the CPCBP demonstrate how emotionally engaging, non-political forms of exchange can foster mutual understanding and goodwill. Second, modern public diplomacy should move away from politicized narratives or grand ideological discourses. Instead, it should prioritize shared human values and emotionally resonant storytelling that transcends cultural and ideological divides. Cultural symbols such as animals, arts, and heritage can serve as neutral yet powerful vehicles for fostering empathy, cross-cultural dialogue, and grassroots-level resonance. Third, effective animal or cultural diplomacy requires contextual sensitivity and strategic customization. One-size-fits-all approaches are unlikely to succeed in diverse international contexts. Nations should develop tailored engagement strategies based on the preferences, attitudes, and socio-cultural context of target audiences. This includes decentralizing public diplomacy efforts by empowering subnational actors—such as local governments, NGOs, and civil society—to participate in storytelling and exchange.

Limitations and future research

While this study provides robust evidence of the CPCBP’s positive impact on international tourism, several limitations should be noted. The analysis focuses on a specific cultural symbol—the giant panda—and a particular set of countries, which may limit the generalizability of the findings. Future research could explore the impacts of other cultural symbols or extend the analysis to different regions and time periods to validate and expand these findings. Although the study successfully addresses endogeneity concerns through instrumental variables, future research could refine these methods further or explore alternative identification strategies. Additionally, longitudinal studies could investigate the long-term sustainability of cultural diplomacy initiatives and their broader socio-economic impacts.

In conclusion, while this study centers on a specific case of animal diplomacy, the theoretical insights and policy implications have relevance for broader efforts to use cultural symbols in international engagement. Future research may explore the comparative effectiveness of different cultural or animal diplomacy programs across countries and contexts, thereby enriching the understanding of soft power in practice.