Introduction

Political, social, environmental, and health challenges have questioned the relevance and sustainability of international student mobility (ISM). Strong critiques towards ISM stem from current global challenges but are also rightfully prevalent among scholars of internationalization. These critiques generally derive from a key feature of ISM: the voluntary movement of some people—often across large distances—in an unequal world. In response to the sustained critique of ISM, this article makes the case that—despite its problems and limitations—ISM can support important public good outcomes and serve broader institutional and higher education sector aims.

Geopolitical shifts and political attitudes have spillover effects on universities and internationalization and on how international students are treated. Following the invasion of Ukraine by Russia in February 2022, the international higher education community has mobilized to support Ukrainian refugees, including scholars and students, and started severing ties with Russian universities. The Russian government has also limited the mobility of its citizens (Delardas et al., 2022). The swift mobilization to support Ukraine and its people has been contrasted with the less welcoming response of the international higher education community during previous geopolitical crises. Australian universities have seen the flow of international students from China greatly impacted by geopolitical tensions between the two countries. Similar tensions between the USA and China have affected internationalization (Golley et al., 2020; Lee, 2020). While public views on immigration are mixed, international students have been previously described in more positive terms and even labeled as “ideal migrants” (Scott et al., 2015) and some countries encourage pathways towards immigration among international students. But in some countries, there has been a backlash against internationalization practices at higher education institutions, due to the belief that domestic students may be crowded out by their international counterparts, despite limited evidence of such effects (Borjas, 2004; Shih, 2017), or due to a perceived link between internationalization and the erosion of domestic culture (Matthews, 2018). All in all, ISM is susceptible to political crossfires.

ISM has also been critiqued on sustainability grounds, in part due to its contribution to global warming and climate change. In 2022, the world was not set on a path to limit global warming to 1.5 °C (2.7 °F). According to the Intergovernmental Panel on Climate Change (2022), unless “immediate and long-term action by governments, businesses, civil society, and individuals at a scale and speed significantly faster than that represented by current trends” (p. 111) were taken, the impacts of climate change may be catastrophic for humankind and the planet we inhabit. Aviation—which is frequently used by international students—has been estimated to be responsible for about 3.5% of global warming and 2.5% of CO2 emissions (Our World in Data, 2020). However, these estimates vary, and recent studies suggest that the effect of aviation on global warming could be higher (Klöwer et al., 2021). Shields (2019) estimates that ISM was responsible for between 14.01 and 38.54 megatons of CO2 in 2014, an increase from between 7.24 and 18.96 megatons in 1999. These figures are not trivial. The lower estimates are comparable to the national annual emissions of countries like Tajikistan or Slovenia (Climate Watch, n.d.). The contribution of ISM to the climate crisis is mediated by ebbs and flows in the number of internationally mobile students (Liu et al., 2022; Shields, 2019). For example, the reduction of student mobility during Covid-19 resulted in a substantial reduction in ISM-related CO2 emissions, leading to calls for online mobilities to more often replace in-person mobilities (Liu et al., 2022). Universities more broadly have their own environmental footprint, with one Canadian study estimating that faculty—often through their own international movement—may have a larger contribution to greenhouse gas emissions per capita than international students (Arsenault et al., 2019). As ISM contributes to one of the largest crises our planet is facing, calls for its reduction are warranted.

During the Covid-19 pandemic, international travel—including that of international students—came under additional scrutiny due to the speed at which the virus has traveled across the world. But the pandemic has been primarily a source of challenge and not problematization for ISM. Travel bans, lockdowns, and discrimination have drastically affected ISM and the experiences of international students. International students, primarily from Asia, have seen an increase in discrimination (Lim et al., 2022; Maleku et al., 2021; Zhai & Du, 2020). The pandemic has also affected mobility choices among international students putting into question the relevance of traditional study abroad destinations (Mok et al., 2021). The prevalence of virtual mobilities and collaborative online international learning (COIL) has (temporarily) increased. The last few years have made visible the complicated experiences of international students in times of crisis and beyond.

In 2020, the average global score of the Democracy Index compiled by the Economist Intelligence Unit was the lowest since the index was created in 2006, in part but not only due to the restrictions on freedom imposed during Covid-19 (The Economist Intelligence Unit, 2020). The democratic decline did not improve by the 2022 edition of the index (The Economist Intelligence Unit, 2023). Some of Europe’s established democracies have seen an increase in dissatisfaction with democracy as a form of government. In 2018, more than 60% of respondents to a Pew survey in Spain, Italy, the UK, Bulgaria, and Greece indicated they are not satisfied with how democracy is working in their country (Wike et al., 2019). With the increase of nationalism and xenophobic attitudes (Bieber, 2018) and skepticism of elites (Wike et al., 2019), also came an increased scrutiny on the role and value of universities that at times spilled into infringement of academic freedom (Craciun & Mihut, 2017). The relationship between these broader political trends and ISM may seem tenuous, but they represent the context in which internationalization takes place and the context that internationalization is aiming to ameliorate (Mihut et al., 2017).

These societal trends challenge and shape the endeavor of internationalization and ISM. But challenges and critiques of ISM also come from within the field and are prevalent among internationalization scholars. One of the leading critiques of ISM is that it may enhance existing inequalities. English-speaking countries receive a disproportionate number of students from other countries. These unequal flows of students allow for center-periphery dynamics to take place (Altbach, 1981), with potential implications for brain drain (de Wit, 2010). These dynamics persist even as evidence of multipolarity is starting to emerge, and more countries are engaging with and benefiting from ISM (Glass & Cruz, 2022). ISM has also been critiqued due to its marketization features. In many countries, including the USA, the UK, and Australia, fees from international students—which are substantially higher than for domestic students—have been used to compensate for decreased public appropriations to higher education (Bound et al., 2020; Miles et al., 2017). Problematic practices, such as the use of agents to recruit international students (Nikula & Kivistö, 2018) and the use of ISM indicators—which reflect existing inequalities within and between higher education systems—to rank universities (Blanco et al., 2021), have also been rightfully critiqued. Calls for inclusivity and social justice rightfully permeate internationalization scholarship (Stein, 2021).

Considering these challenges and sustained critiques towards ISM, this paper aims to make the case that ISM continues to be relevant, may help promote public good (Marginson, 2011), and may help ameliorate some of the very societal challenges we have seen turn the world upside down. The contribution of this paper is both discursive and empirical.

Empirically, this paper has two contributions. First, the paper tests the associations between ISM and public good markers, as described later. Second, this paper juxtaposes the public good outcomes of international graduates with not only domestic graduates—as traditionally done in the academic literature—but also with those of migrant graduates. Migrant students and graduates—who have had some experience with the domestic education system prior to entering higher education—are an under-used population in internationalization activities and their cultural contributions are not equally celebrated. The inclusion of migrant graduates in this paper signals a broader, more comprehensive view on internationalization. It also has a practical aim, as it helps identify areas of additional support needed for different groups of students (Mihut, 2022).

Discursively, the paper also has two contributions. First, it brings into focus the public good purpose of ISM. The purpose of ISM may be context-dependent and debatable, and the specific vision proposed in this article may be contested. But this article aims to recalibrate recent critiques and challenges to ISM by serving as a reminder of its importance and potential contribution to the public good. Second, it spotlights migrant students and graduates in internationalization debates as a way to broaden the conception of what internationalization resources are available on campuses (Mihut, 2022). Using international graduates as a comparison group to understand public good outcomes of higher education is one of the ways through which ISM can become more useful to the sector and can help enhance the public good mission of higher education (Mihut, 2019; Mihut, 2022).

This paper asks the following questions:

  1. 1.

    Is international student mobility associated with trust in others, political participation, higher self-reported levels of competency, and happiness?

  2. 2.

    Do international graduates, migrant graduates, and domestic graduates differ in their trust in others, political participation, self-reported competency levels, and happiness?

The public good, democratization, and human development

Throughout this paper, public good is understood as a normative and collective good and operationalized to encompass markets of democratization and markers of human development. This approach draws on Marginson’s (2011) concept of public good (singular) and stands in contrast to an economic view of public goods as non-rivalrous and non-excludable (Samuelson, 1954). In (economics of) higher education, the concept of the public good is often juxtaposed to that of private goods or private returns (Marginson, 2007). Such discussions aim to quantify if the individual or the society benefits primarily from higher education. The results of these quantifications are commonly used to argue who should pay for higher education, the state or the individual, but also to discuss the role of state funding more broadly in higher education and the role of taxation in this endeavor (Marginson, 2017). This economic framework has also been applied to understand other policy decisions regarding higher education and even international higher education (Cai & Kivistö, 2013; Miles et al., 2017; Oleksiyenko et al., 2013). Economic public good calculations have prompted countries to justify different and higher tuition costs for international students than domestic students in a variety of contexts.

The economic view of public good is not the conception of public good used by this study. Instead, the focus here is on what Marginson (2011) calls public good (singular). Marginson defines public good (singular) as bringing universities into a “larger process of democratization and human development” (p. 418). In this paper, I suggest that the same can be said about ISM.

Many studies that look at democratization see it as an ongoing process (Grugel & Bishop, 2013) and are particularly interested in how countries make the transition from non-democratic regimes to democratic regimes and/or vice versa (Huntington, 1993; Skaaning, 2020). In this paper, democratization is not understood as a transition process, in part because the data used here was collected from countries (Austria, Czech Republic, Croatia, Germany, Greece, Lithuania, Malta, and Norway) that are regarded as democracies (The Economist Intelligence Unit, 2023). Instead, the definition put forward by the journal “Democratization” is more suited to the purposes of this study. Democratization is a phenomenon in which “democratic norms, institutions and practices evolve and are disseminated or retracted” (Democratization, n.d., para. 1). As a concept, democracy is highly contested, but aspects such as pluralism, freedom, fragmentation, and change of power are main features. One of the key characteristics of democracies is the availability of forms of political participation, in which individuals are free to engage in the political process. In fact, political participation is often used as an indicator of democracy (The Economist Intelligence Unit, 2023). What constitutes political participation has evolved, from a narrow focus on voting in the 1940s and 1950s to include aspects such as boycotting and protesting more recently (van Deth, 2016). Often associated with social capital theory, trust has been shown to promote democracy (Zmerli & Newton, 2008). In the context of democracy studies, trust refers to both (1) political trust, or how individuals view institutions, governments, and representatives, and (2) social trust, or how individuals view other individuals (Warren, 2018; Zmerli & Newton, 2008). In this paper, I use self-reported levels of social trust and political participation as operationalizations of democratization.

Marginson does not offer a definition of human development. But for the purposes of this paper, human development is understood as the enrichment of individuals beyond economic aspects. Self-reported competency levels and happiness are used as operationalizations of human development.

The measures of democratization and human development used in this paper are imperfect as they rely on measures of public good markers at the individual level and are based on self-reported survey data, but they allow observing empirically associations between ISM and broader aspects that contribute to democracy building and human development.

The relationship between ISM and the public good is looked at in two ways in this paper. First, the paper looks at the association between study abroadFootnote 1 as a form of ISM, on one hand, and the outcomes associated with public good considered in this study, on the other hand. Second, the paper investigates how the same outcomes vary between populations with different exposures to international experiences. Specifically, the paper looks at variations between international graduates, migrant graduates, and domestic graduates.

Literature review

Previous studies have reflected on ISM and public good jointly, but often such studies are not accompanied by empirical analyses that exemplify the contribution of ISM to the public good. Instead, scholarship has investigated how universities make use of public good narratives when marketing themselves to international students (Rhoades et al., 2019), whether stakeholders view public good to be associated with internationalization policy and practice (Huang & Horiuchi, 2020; Tian & Liu, 2021), and if governments make use of public or private good logics to make policy decisions about ISM (Oleksiyenko et al., 2013; Tian & Liu, 2021). In contrast, the literature investigating specifically the association between education and democratic and human development markers is more developed.

Multiple studies have been conducted on the importance of education for democracy and democratization. These studies have shown a relationship between the two (see Bobba & Coviello, 2007; Castelló-Climent, 2008; Dewey, 1903). Previous studies have also documented democratic involvement among university students (Van de Velde, 2022). Yet fewer studies investigate the link between internationalization and democracy or democratization. The existing evidence suggests that international students contribute to democracy building in their native countries if they studied in a democratic country (Chankseliani, 2018; Spilimbergo, 2009). This phenomenon has been visible in post-soviet countries, with countries that had higher proportions of students studying in Europe or the USA having achieved higher levels of democratic development (Chankseliani, 2018). Previous studies have also investigated qualitatively how the experiences of students from non-democratic countries studying in democratic countries may shape views of citizenship (Zhang, 2018).

The relationship between study abroad and trust in others has not received much attention in the academic literature; however, studies have previously focused on the association between study abroad and the attitudes of international students towards immigration, after accounting for socio-economic characteristics (Öz & Gök, 2022) and social engagement (Souto-Otero et al., 2019). Previous studies have also documented decreased xenophobic and nationalistic attitudes among populations with higher levels of education, but these studies do not generally include comparisons between students with and students without international exposure (Coenders & Scheepers, 2003; Hjerm, 2001). Studies have not investigated extensively the association between study abroad and political participation at the individual level. Previous analysis using Eurograduate data suggests that participation in an international mobility and other voluntary activities is positively correlated with political participation as well as having higher levels of trust in others (Meng et al., 2020).

The link between ISM and markers of human development is more commonly examined in the academic literature. Self-reported competency levels, particularly for foreign language skills, were found to be higher among internationally mobile students (Howard, 2019; Martinsen et al., 2010; Souto-Otero et al., 2019). International students have also been documented to have stronger academic results than domestic peers, but these results vary between different populations of international students (Rienties et al., 2012). Previous studies have focused on understanding the factors that motivate students to pursue ISM and have identified pursuing happiness as one of the motivators (Waters et al., 2011). However, investigations about how long-term happiness levels differ between international students and other students are lacking.

Few previous studies have investigated how outcomes vary between international and migrant students and how their outcomes compare to their domestic peers. A previous study based in the US context has found that international students are more likely to complete their bachelor’s degree and they reported higher levels of academic and socio-cultural engagement than their migrant peers (Lee et al., 2021). Anderson et al. (2022) contrast the experiences of international and refugee background women in New Zealand and Bangladesh, highlighting how choice and autonomy are viewed by these women. Using Eurostudent data from Ireland, one study has highlighted that migrant students report lower levels of preparation for the domestic labor market than both domestic and international students (Mihut, 2022).

Cumulatively, previous studies suggest a positive link between ISM and trust, political participation, and self-reported levels of competency. Previous evidence on the relationship between ISM and happiness is lacking. This study builds on existing literature by looking at how these outcomes manifest themselves among graduates rather than students and by discussing the four outcomes in the broader framework of public good.

Methodology

The paper makes use of Eurograduate 2018 survey pilot anonymized data.Footnote 2 The data was collected from recent graduates from Austria, Czech Republic, Croatia, Germany, Greece, Lithuania, Malta, and Norway between October 2018 and February 2019. The study targeted two cohorts of graduates: (1) the graduating cohort of 2012/2013, and (2) the graduating cohort of 2016/2017. For the graduating cohort of 2012/2013, data was collected 5 years after graduation, while for the 2016/2017 cohort, data was collected 1 year after graduation. The Eurograduate pilot survey gathered data from ISCED 6 (bachelor’s) and ISCED 7 (master’s) graduates from public and private higher education institutions in each country and who attended either universities or other types of post-secondary institutions. Data was gathered from graduates of the targeted cohorts regardless of nationality. For the 2012/2013 cohort, response rates varied between 5.4% for Lithuania and 63% for Germany. For the 2016/2017 cohort, the response rate varied between 9.3% for Lithuania and 21.9% for Norway. Response rates could not be calculated for Greece. The overall response rate for the survey was 12% (Mühleck et al., 2021). This response rate is low, but in line with other large-scale surveys. Eurograduate cautions that the data collected through the pilot survey is not comparable across countries, as definitions of graduates as well as sampling strategies differed between countries (Meng et al., 2020). However, Eurograduate allows for the same definition of study abroad, as well as migrant, international, and domestic students to be used across countries. Data was weighted using population-level data (e.g., ISCED level, ISCED broad field, type of higher education institution, and sex) to ensure its representativeness for the student population in each country. Eurograduate used a “raking procedure” for weighing, which allows comparability between countries (see Meng et al., 2020, p. 32 for additional detail). While the primary goal of the Eurograduate survey was to gather data on the transition from university to the labor market, the survey gathered data on other aspects including demographic characteristics, skills, mobility, democratic values, and citizenship.Footnote 3

For the purposes of this paper, domestic graduates are defined as those graduates who have no migration background and received their secondary school degree from the country of reference. Migrant graduates include both first- and second-generation migrants. First-generation migrants were not born in the country of reference but completed a secondary degree in the country of reference. Second-generation migrants had one or both parents born in a different country than their country of reference. In most but not all cases, these would have completed a secondary degree in the country of reference. International students were not born in the country of reference and have not completed a secondary degree in the country of reference (for a similar definition, see Mihut, 2022 and Hauschildt et al., 2018). Using this definition, 80% of survey participants were identified as domestic graduates (n = 12,207), 15% were identified as migrant graduates (n = 2320), and 5% were identified as international graduates (n = 686). The definitions are also exemplified in Table 1. Study abroad participation was similar across the three groups of graduates, with 13% of domestic graduates, 13% of migrant graduates, and 16% of international graduates having studied abroad during their degrees.

Table 1 Defining domestic, migrant, and international graduates
Full size table

The independent variables included in the models capture personal, socio-economic, and institutional characteristics that have been previously found to relate to one or more of the outcome variables considered. These variables also follow a broad ecological approach (Bronfenbrenner & Morris, 2006). Personal and socio-economic independent variables included in this paper are (1) the sex of participants (female and male only); (2) whether participants were the first in their family to attend university, had one parent who attended university, or both parents attended university; and (3) age at graduation. A dummy variable capturing the Eurograduate cohort was also included. Institutional variables included are (1) language of instruction (studied in the official language of the reference country or not); (2) mode of study (full-time or part-time); (3) type of institution attended (university or non-university); (4) level of study (masters or bachelors); and (5) whether respondents participated in study abroad (as a proxy of international student mobility).

Previous studies illustrate the relevance of these variables for the outcomes considered in this paper. Previous studies have shown that participant’s sex matters for trust (Dorrough & Glöckner, 2019; Maddux & Brewer, 2005), political participation (Liu, 2022), and happiness (Stevenson & Wolfers, 2009), but these associations are mixed and context-dependent. Previous studies have generally not explored relationships between students who are first in their family to attend university and markers of democratization. But previous analysis of Eurograduate data suggests that first-generation students have lower levels of trust in others than their peers and similar levels of happiness as their peers. There is mixed evidence regarding the association between first-generation status and self-assessed competency levels (Meng et al., 2020). Age too has been previously found to be associated with political participation (Melo & Stockemer, 2014) and happiness (Wang & Wong, 2014). While the association between studying part-time or full-time and democratic markers or happiness has not been documented, previous evidence suggests a relationship between flexibility (for example in the workplace) and happiness (Okulicz-Kozaryn & Golden, 2018).

The breath of variables included in the analyses of this paper was limited by the availability of data in the anonymized version of the Eurograduate 2018 dataset. Data availability also impacted on the quality of a few measures included in this paper. Important personal and context characteristics, such as race, ethnicity (Leighley & Vedlitz, 1999), social class, national context of origin (Gallego, 2007), political socialization (Sani & Quaranta, 2015), and geographic mobility/immobility within the domestic country (Lee et al., 2018), are missing from this analysis and other unobserved characteristics may affect the associations discussed in this paper. An exploration of additional unobserved characteristics that may matter here is beyond the purpose of this paper. But one key limitation of the Eurograduate dataset is the absence of pre- and post-study abroad measures. This means that the analysis in this paper cannot control for any selection effects associated with who chooses to study abroad. This limitation is particularly important as previous studies have shown that participants who choose to study abroad already have higher social and cultural openness (Brandenburg et al., 2014; Souto-Otero et al., 2019).

Four outcome variables are considered in this paper. Markers of democratization are operationalized as (1) trust in others and (2) political participation. Markers of human development are operationalized as (3) self-reported competency levels and (4) happiness. As part of the Eurograduate survey, participants were asked “Would you say that most people can be trusted or that you can’t be too careful?” Despite some limitations, this question has been frequently used to measure social trust and was devised in 1948 by Noelle-Neumann (Delhey et al., 2011). In this study, self-reported levels of trust in others were coded as a dummy variable (you can’t be too careful = 0; neither/nor = 0, and most people can be trusted = 1).

To measure political participation, survey participants were asked “There are different ways of trying to improve things in your country or help prevent things from going wrong. During the last 12 months, have you done any of the following?”, with “yes” and “no” as answer options. Nine items were included in this question: (1) contacted a politician, government or local government official; (2) donated to or participated in a political party or pressure group; (3) worked in another politics-related association; (4) worn or displayed a campaign badge/sticker; (5) signed a petition; (6) taken part in a public demonstration; (7) boycotted certain products; (8) posted or shared anything about politics online, for example on blogs, via email or on social media such as Facebook or Twitter; (9) volunteered for a not-for-profit or charitable organization. The items included in questions aimed to measure political participation vary slightly between studies and instruments (van Deth, 2014). Eurograduate aimed to allow comparability with the European Value Survey and the European Social Survey (Eurograduate, 2018). In this paper, political participation is treated as a dummy variable (engaged in at least one form of political participation = 1; else = 0).

Self-reported competency levels were measured using a Likert scale from 1 (low) to 5 (high). Participants self-reported their current level of competency for (1) own field-specific skills, (2) communication skills, (3) team-working skills, (4) foreign language skills, (5) learning skills, (6) planning and organizing skills, (7) customer handling skills, (8) problem-solving skills, and (9) advanced ICT skills. The mean value across these items was used to operationalize self-reported competency levels. Participants were also asked to rate their happiness levels on a scale from 0 to 10. Responses to this question were used to develop a dummy variable for happiness (8, 9, 10=1, else=0). No missing data imputations were conducted for this analysis. Ordinary least square (OLS) regression using a linear functional form (self-reported competency levels) and logistic regressions (trust, political participation, and happiness) were used as methodological approaches in this paper. No interactions were included in these models. No transformations were performed in the analyses.

Logistic regressions are conducted for the political participation, trust, and happiness outcome variables because OLS assumptions for these models (and variables) were not met. These outcome variables were not normally distributed and homoscedasticity assumptions were not met. Normality, multicollinearity, and homoscedasticity assumptions were met for the self-reported competency level variable. The analysis was conducted in Stata 17.

Results

This paper aims to observe if ISM is associated with democracy markers and human development markers as operationalizations of public good (singular) (Marginson, 2011). Two distinct aspects of ISM are considered in this analysis: (1) participating in study abroad, and (2) how outcomes differ between international, migrant, and domestic graduates. Table 2 includes the results of the inferential analyses conducted for this paper. These results are based on self-reported data from the Eurograduate pilot survey. Respondent selection bias is an important limitation of such surveys, and this may affect the reliability of the results presented in this paper.

Table 2 Regression results
Full size table

After accounting for selected personal and context characteristics, as well as cohort effects, participating in study abroad was associated with increased trust in others, increased political participation, increased self-reported levels of competency, and increased likelihood of reporting high levels of happiness. These results underline the potential contribution of study abroad to public good outcomes that may enhance democracy and human development. Study abroad was the only variable that registered positive and statistically significant associations with each of the four outcome variables considered in this paper. However, this association is weak. Importantly, existing variables explain less than 4% of the variability in the outcome variables, and many potentially relevant variables were not included in these models due to data availability. See Table 2 for regression results.

Increased public good outcomes were more common but not universal among study abroad participants. Among graduates who did not study abroad, 37% indicated that most people can be trusted. Graduates who studied abroad were 1.5 times more likely to indicate that most people can be trusted (55%). Among graduates who did not study abroad, 65% indicated they engaged in at least one form of political participation. In comparison, 71% of their peers who had studied abroad engaged in at least one form of political participation. Graduates who participated in study abroad had only slightly higher average self-reported competency mean scores (M = 3.79) than their peers (M = 3.73), and the difference in scores was primarily due to higher self-reported language skills among graduates who participated in study abroad. If 57% of those who did not participate in study abroad indicated high levels of happiness, the equivalent figure was 61% among graduates who have participated in study abroad (see Figure 1).

Fig. 1
figure 1

Trust, political participation, self-reported skills, and happiness by study abroad

Full size image

Outcomes varied among the three groups of students considered in this paper: international graduates, domestic graduates, and migrant graduates (see Table 1 for definition and Figure 2 for descriptive results). International graduates reported statistically significantly higher levels of competency (M = 3.83) and happiness (60%), yet they were also less likely to engage in any form of political participation (63%) when compared to their domestic peers. In contrast, migrant graduates were statistically significantly less likely to trust others (35%) and reported lower competency levels (M = 3.72) but were statistically significantly more likely to have engaged in at least one form of political participation (67%) than their domestic peers. Among domestic graduates, 41% indicated that most people can be trusted, 66% engaged with at least one form of political participation, and 58% reported high levels of happiness. The differences between the three groups of graduates (Figure 2) were lower than the differences between graduates who have studied abroad and those who have not studied abroad (Figure 1).

Fig. 2
figure 2

Trust, political participation, and self-reported skills among domestic, migrant, and international graduates

Full size image

Across models, no statistically significant differences were registered between female and male participants regarding trust in others and self-reported competency levels. Instead, male graduates were less likely than female participants to engage with at least one form of political participation and they were less likely to report high levels of happiness than female graduates. Both graduates who were the first in their family to attend university and graduates who had only one parent that previously attended a university showed lower levels of trust in others than their peers from a family in which both parents received a higher education degree. Graduates who were the first in their family to attend higher education reported higher levels of competency (M = 3.77) than their peers from families where both parents attended university (M = 3.74). Participants who were 24 or younger were less likely to trust others, less likely to participate politically, reported lower competency levels, and more likely to report high levels of happiness than their older peers. Differences were also noted between the two cohorts which participated in the Eurograduate 2018 pilot. Graduates who completed their study in 2016–2017 and were 1 year out from their studies were less likely to trust others, reported lower competency levels, and were less likely to report high levels of happiness than 2012–2013 graduates, who were surveyed 5 years after graduation.

Trust in others was higher among graduates who had studied full-time than those who studied part-time. The same group of graduates reported lower competency levels and were less likely to be very happy than their peers who studied part-time. Graduates from a university were more likely to participate politically than their peers attending other types of post-secondary education institutions. While master’s graduates reported higher competency levels and were more likely to report higher levels of happiness than bachelor’s graduates, they were also less likely to engage politically.

Discussion and conclusion

This paper shows that studying abroad may represent a protective factor for democratization and human development markers, thus potentially helping address some of the political and social challenges currently faced at national, regional, and global levels. Yet findings from this paper also suggest that study abroad is a contributory, not a necessary or a sufficient factor in this relationship. Graduates who had studied abroad were more likely to exhibit the public good outcomes considered in this paper after accounting for other personal, social, and institutional characteristics, yet some study abroad participants persisted in not trusting others, were not engaged politically, and were not particularly happy. Another difficulty in assessing the relationship between study abroad and public good outcomes is posed by the self-selection of those who choose to pursue various forms of mobility. This paper is unable to assess if study abroad may matter for public good outcomes due to the experiences that accompany study abroad, if study abroad serves as a proxy for other unobserved characteristics, or if people who are more trusting of others, more engaged politically, and happier are more likely to choose to study abroad. This is particularly relevant as previous studies that compare pre-mobility outcomes with post-mobility outcomes have shown that study abroad participants have higher social and cultural openness before their study abroad (Brandenburg et al., 2014; Souto-Otero et al., 2019). The variations in outcomes between international graduates and migrant graduates may hint at some of these selection effect dynamics. As outcomes vary between international and migrant graduates, perhaps choice and self-selection matter more than exposure to international experiences. Or perhaps the experiences of international graduates and migrant graduates are very distinct (Lee et al., 2021). Additional analyses are also needed to understand why the selected outcomes vary between those who studied abroad and international graduates.

Overall, the findings of this paper are consistent with previous research. This paper finds that study abroad is linked to increased trust in others and political participation (Meng et al., 2020; Öz & Gök, 2022). It also confirms previous findings of higher self-reported competency levels among graduates who studied abroad, particularly regarding foreign language skills (Howard, 2019; Martinsen et al., 2010). However, self-selection effects may explain these associations. Further studies can use longitudinal and quasi-experimental research approaches to investigate the relationship between trust in others, political participation, self-reported competency levels, and happiness, on one hand, and study abroad, on the other hand.

Previous studies which contrasted the experiences and outcomes of international students with those of migrant students have shown that international students generally fare better than their migrant peers (Lee et al., 2021; Mihut, 2022). This study also shows a complex picture on how outcomes vary between domestic, migrant, and international graduates. Migrant graduates indicate lower likelihoods of trusting others and report lower competency levels than both domestic and international graduates. This finding has implications for institutional practice and policy. Together with evidence from previous studies, this finding suggests that higher education institutions should offer additional support to migrant students.

Not captured by this study are how outcomes may vary among sub-populations of study abroad students and graduates. Previous research has highlighted that international students from less democratic contexts may be shaped by their study abroad experience in a democratic country (Spilimbergo, 2009; Zhang, 2018). Further research can investigate if differences in political regimes between home and host countries impact on political participation and trust. As a limitation, this study treats study abroad as a homogenous experience. Previous studies have shown how complex and heterogenous the international student experience can be. Outcomes and experiences can differ between international students from different regions and countries of origin (Darmody et al., 2022; Finn et al., 2021; Rienties et al., 2012) and further studies can further disaggregate between groups of international students.

While the analysis in this paper discusses the potential contribution of ISM to some of the societal challenges discussed in the introduction of the paper, two of these important challenges to ISM have received no empirical focus in this article: (1) the impact that ISM has on the environment and (2) inequalities associated with ISM. Value judgements and diverging priorities may still make us decide that study abroad and ISM may need to change or diminish in the face of crises. Here, an increase in use of virtual exchanges and initiatives like COIL may potentially help mitigate some of the effects of ISM on the environment and may bring this opportunity to more students that would traditionally be excluded from ISM. Yet evidence on how outcomes vary between students and graduates who benefit from physical mobility and those who participate in virtual mobility is only now starting to emerge. To better support policy advice in this area, further studies should investigate how the broader public good outcomes associated with ISM can be replicated in virtual environments as well as other non-mobility related internationalization activities.

Studies that look at the value of education and ISM often use either an empirical approach that is not embedded in normative values or rely on aspirational statements without evidence. Operationalizing the concept of public good in the context of ISM is contentious. One of the contributions of this paper is to embed the empirical analysis put forward in a normative framework that can be used by other studies that theorize and investigate the value and relevance of ISM. Framing the value of ISM using the broader concept of the public good (singular) can help us have constructive debates that can both improve and perhaps defend the value of ISM in light of the many challenges it faces.

The evidence in this paper supports findings from previous studies about the link between study abroad and democratization and human development markers. Taken together with previous evidence, this paper offers a hopeful take on how study abroad may contribute towards public good outcomes in higher education. As political, social, environmental, and health challenges raise questions about ISM and as scholars of internationalization rightfully critique multiple aspects of this endeavor, it is important to remember the benefits of studying abroad. Marginal positive contributions from ISM to democratization and human development markets can perhaps contribute to the public good aims of higher education institutions and internationalization in a world turned upside down.