Abstract
Western education is credited with producing democratic reformers from non-Western countries—and blamed for training dictators, elitist technocrats, and antiliberal leaders. Why do similar programs generate such divergent political trajectories? Drawing on an international, multidisciplinary literature, this article critically analyzes 94 studies to explore when, how, and for whom Western education influences democratic attitudes and outcomes in non-Western countries. It develops an interdisciplinary framework that pinpoints key theoretical and methodological themes explaining how Western educational programs can foster democratic engagement, entrench illiberal rule, or have limited political impact. The analysis highlights underexplored contextual factors shaping trainees’ experiences, and proposes methodological strategies—individual-level, longitudinal, comparative designs embedded in home-country contexts—to move the field beyond the currently thin empirical debate. Against a backdrop of rising East–West tensions, the article aims to inform realistic judgments about the prodemocratic potential of Western educational programs—and the risks of relying on them uncritically.
Keywords
“The ideas and values that America exports in the minds of more than half a million foreign students who study every year in American universities and then return to their home countries . . . tend to reach elites with power.” —Joseph Nye, Soft Power (2004, p. 13) “I lived in America. I love America . . . We are not Communists anymore. We have changed. . . . You don’t have to be afraid of us anymore.” —Margarita Simonyan, Vladimir Putin’s chief propagandist, former exchange student in the U.S., in a 2017 interview with NBC News on Russia’s meddling in 2016 U.S. elections
Serious attention to the question of Western education’s potential to promote—or hinder—democracy in non-Western societies is overdue. Tensions between the Global North and South, East and West have risen to alarming levels worldwide. While often celebrated as bedrocks of human progress since the European Enlightenment, the models of scientific inquiry, education, economic development, and social organization exported to the rest of the world by Western countries have also been increasingly accused of parochial universalism, Eurocentrism, or (neo)colonialism (Brohman, 1995; Said, 1979; Santos, 2015; Tikly, 2004; Wallerstein, 1997).
Meanwhile, prodemocratic actors across the world are anxiously watching the global onslaught of democratic backsliding, populism, and technologically advanced authoritarianism (Bermeo, 2016; Lührmann & Lindberg, 2019; M. Wang, 2021). Western societies show no immunity to such challenges, but in many non-Western countries, the lack of democracy has also often bottlenecked addressing other major societal challenges (Berman, 2021; Carothers & de Gramont, 2013; Gerring et al., 2022). The record of traditional external democracy assistance in these nations is modest (Avdeenko & Gilligan, 2015; Beath et al., 2013;Bush, 2016; Carothers, 2015; Carothers & de Gramont, 2013; Knack, 2004), and current trends in global politics further constrain the effectiveness of such efforts—positioning Western education as one of the few potentially viable strategies for promoting democracy and addressing these societal challenges. But what do we really know about such programs’ effectiveness in promoting democratic changes?
The literature on the impact of Western education on democracy in non-Western countries is sharply divided. On one hand, many scholars, politicians, and funders contend that training non-Western citizens in Western institutions—through formal education, public diplomacy programs, and professional training schemes—fosters democratic development back home (Atkinson, 2010; Chankseliani, 2018; Gift & Krcmaric, 2016; Mercier, 2016; Ruby & Gibler, 2010; Spilimbergo, 2009). On the other hand, some argue that Western education may actually hinder democratic change. They point to numerous examples of Western-trained individuals from Latin America to East Asia who became dictators, elitist technocrats, infamous soldiers, or antiliberal Islamists (Dooley, 2015; Kapur, 2014; Monbiot, 2001; Newton, 2002; Sordi, 2018). Furthermore, some question the relevance of Western education for political change in non-Western countries altogether (Dzotsenidze, 2021; Peachey, 2017; Silva, 1991; I. Wilson, 2016). Given these starkly contrasting views, what conclusions can we draw?
Conceptualizing “the West” and “Western Education”
In this article, I retain the terms “the West” and “Western education” primarily because they are the dominant conceptual labels used in the literature I review to describe education obtained in a set of countries commonly grouped as Western democracies (listed in the Methods section). I treat the West here as a cluster of countries rather than as a metaphorical positional concept akin to “Global North/South.” The term is prevalent in academic and policy literatures closely related to this article’s focus—for example, in international relations (“Western-led international order”), democratization (“Western democracy assistance”), development debates (“Western donors”), human rights (“Western human rights norms”), migration (“migrant integration in Western countries”), and in public discourse in many settings beyond Western democracies (“resisting Western imperialism”). It is geographic insofar as it refers to specific countries commonly understood as Western. At the same time, the notion of the West is to some extent meaningful as a heuristic that reflects partly shared cultural legacies, historical developments, and political institutions (particularly longer and more sustained experiences with liberal democracy), often linked to origins in the European Enlightenment. I situate these links within the broader critical debates on Eurocentrism and (neo)colonialism introduced in the opening paragraph.
The term “Western education” refers to education systems and practices in these Western countries. I use it as a historically contingent heuristic, an operationalization that appears reasonable given how the existing literature groups these cases. This heuristic is grounded in a set of features that many studies associate with these systems: pedagogical practices that emphasize critical discussion, argumentative writing, and student participation; comparatively high levels of institutional autonomy and academic freedom; and curricular structures that expose students to the social sciences and humanities and, in some instances, liberal-arts-style general education. Much of the literature I review assumes that such environments convey liberal-democratic norms and civic skills, and that studying in them may shape political attitudes and practices among students from outside Western democracies. However, one of the critiques this article advances is that the widespread use of this heuristic risks masking important variation, reproducing a simple West/non-West divide that I analyze as a construct of the literature rather than endorsing, and sliding toward an implication of civilizational essence.
Objectives
Given the starkly contrasting assessments of Western education’s impact, the conceptual ambiguities discussed previously, and the fact that any new primary study would be filtered through researchers’ explicit or implicit theoretical predispositions, methodological choices, and data limitations, another conventional primary study is unlikely to advance the debate beyond this stalemate. Instead, this article proposes a promising and timely alternative: a critical integrative review.
My first objective is to synthesize existing knowledge critically and comprehensively, documenting advancements in our understanding of the sociopolitical impacts on non-Western societies of having their citizens educated in Western training programs. The expanding multidisciplinary body of literature on this topic shares many ideas and provides diverse insights. However, contributions from various disciplines and research programs are often overlooked, leaving the potential benefits of synergy untapped and causing knowledge accumulation to proceed slowly and laboriously. Therefore, my second objective is to develop an up-to-date, systematic, and interdisciplinary framework that allows these diverse voices to engage more productively with one another, fostering much-needed cross-pollination at a critical time.
The article draws on research covering different continents, countries, groups, and programs. Given the salience of optimistic accounts in mainstream Western policy and academic circles, the following analysis is often framed as a conversation with such arguments. But pessimistic and skeptical accounts receive similar scrutiny. The focus of this review is on the impacts of Western education programs on political attitudes, behavior, and institutions in non-Western countries. However, many of the theoretical and methodological challenges identified here apply to research programs in related fields, such as diverse pro-developmental effects of Western education and other impacts of international professional training, public diplomacy, and citizen diplomacy.
Contributions
Alongside offering an up-to-date interdisciplinary synthesis, this article identifies key theoretical and methodological themes to advance knowledge in the field. The reviewed studies draw on a range of theoretical frameworks. “Optimistic” perspectives on the impact of Western education programs often build on modernization theory, political participation theories emphasizing education and liberal socialization, soft power, linkage and leverage theory, and favorable views of democratic norm diffusion. In contrast, “pessimist” perspectives tend to draw on critical alternatives to modernization theory, constructivist critiques of norm diffusion, and theories of authoritarian stability. By bringing these divergent theoretical perspectives into conversation and evaluating their explanatory strength in this domain—while drawing on a variety of complementary approaches and sources of evidence—this article highlights consequential yet underexplored contextual factors that shape trainees’ experiences before, during, and after their Western education, particularly the influence of social structures, systemic filters, and inequalities on individual incentives, agency, and actions.
This review identifies key theoretical challenges across differing accounts, underscoring the need to move beyond broad conceptions of Western education toward a more systematic consideration of the critical diversity among Western destinations, study programs, and trainees. It stresses the importance of systematically examining the subtle yet consequential assumptions that underpin various arguments. Furthermore, it highlights the value of incorporating individuals’ social and cultural capital, investigating the actual sociopolitical engagement of alumni, and integrating home-country contexts into the analysis.
Accordingly, the article identifies corresponding methodological challenges and proposes strategies to address them. These include generating systematic evidence at the individual level, explicitly comparing Western-trained individuals with appropriate control groups, adopting longitudinal designs, systematically comparing home-country contexts, and improving both methods and measurement tools.
Despite the substantial benefits it offers, a comprehensive, critical, and interdisciplinary synthesis on this topic has not yet been undertaken. By weaving together findings from different communities of practice and deriving new ideas, this article contributes to advancing relevant debates mainly in three fields. First, it draws insights from and speaks directly to international education studies, focusing on the experiences of international trainees. The findings highlight serious implications, examining often-overlooked aspects of international education programs in mainstream Western economic and policy debates, especially their sociopolitical costs and benefits for non-Western countries.
Additionally, the article suggests avenues to deepen the theoretical and methodological frameworks used in political science, particularly in its subdiscipline of international relations, to study the domestic political impacts of Western education in non-Western societies, international norm diffusion, and East–West interactions, focusing on international trainees as a significant yet frequently overlooked group of actors. Finally, as international trainees also represent a prominent group of migrants, the article engages with migration studies debates on “social remittances,” viewing migrants as transnational agents of sociopolitical change in their home countries.
Policy-wise, this article aims to foster a debate on how to fully realize the prodemocratic potential of Western educational programs, while avoiding the extremes of West-centrism and anti-Westernism. It highlights avenues for thoughtfulness in offering Western educational opportunities and amplifies the call for improving evaluation practices through a deeper, more systematic examination of complex outcome-related questions (Mawer, 2014). The article explores ideas to ensure that Western training opportunities effectively foster democracy without inadvertently contributing to the reproduction of sociopolitical inequities in non-Western societies. Finally, recognizing that citizens in both Western and non-Western countries often fund international study programs and hold high expectations for their transformative impact, it aims to deepen public understanding of these intricate issues.
Background: Western Education and the Transfer of Ideas
While foreign education is an old phenomenon and the history of bursaries to study abroad can be traced back to Ancient Egypt (Arndt, 2005), only in the last 70 years has foreign training reached remarkable salience and become one of the main forms of international mobility. While in the 1950s there were around 50,000 people studying abroad, this number reached 2 million in 2000 and doubled to over 4.1 million in 2013—in that year, out of 100 students enrolled in tertiary education, two studied abroad (Spilimbergo, 2009; UNESCO, 2020). But it tripled further to 6.1 million in 2019 (UNESCO, 2020). Since the Cold War, various public diplomacy programs, international professional trainings, and scientific and cultural exchanges also proliferated. Financial support for foreign training has also grown spectacularly in number, global scope, and the size of its cohorts (Perna et al., 2014).
Advanced Western democracies have led both as hosts of international trainees and as sponsors. In 2020, six Western democracies hosted around half of all international students: the United States (16%), the United Kingdom (9%), Australia (8%), Germany (6%), Canada (5%), and France (4%) (UNESCO, 2020). By 2010s, the U.S. government had given more than 158,000 Fulbright scholarships alone to foreign individuals to study in the United States (Spilimbergo, 2009), the UK’s Chevening program had a global network of alumni exceeding 40,000 (Chevening Program, 2014), and the German Academic Exchange Service (DAAD) funded more than 50,000 foreigners in 2015 alone (DAAD, 2016).
In international education, most developing countries are suppliers of students and thus “net importers of higher education” (Adnett, 2010; UNESCO, 2020). In 2020, China, India, Viet Nam, and Turkmenistan had the highest net import rate with −863,366; −123,913; −68,079; and −66,787, respectively. In contrast, the United States, the United Kingdom, Australia, Canada, and Germany had net inflows of 847,648; 510,803; 444,537; 272,001; and 245,205, respectively. In terms of destinations, in 2020 majorities of students from largest non-Western countries studied in North America and Western Europe: China (61%), India (64%), Indonesia (34%), Pakistan (45%), Brazil (62%), Nigeria (57%), Bangladesh (51%), and Russia (58%). These numbers are even higher for several other major non-Western countries: Algeria (84%), Morocco (75%), Saudi Arabia (74%), Argentina (73%), and Turkey (69%) (UNESCO, 2020). While lagging behind in scholarships, non-Western countries are increasingly offering more resources to fund their nationals’ studies abroad (Perna et al., 2014).
Foreign sponsors, sending communities, and many Western trainees themselves hold high expectations regarding the potential of West-bound international scholars, students, and professionals to transform their societies of origin (Diamond, 1994; Perna et al., 2014; E. C. Wilson & Bonilla, 1955; Ye, 2002). Despite being relative minorities in their countries of origin, foreign-trained individuals are seen as elites with a strong potential to lead and to transfer norms, knowledge, and practices to their home countries (Kapur, 2014), reduce extreme poverty long term (Kwak & Chankseliani, 2024) or improve home-host country relations through everyday diplomacy (Atabaş & Köse, 2024). Given the perceived benefits of international and particularly Western training, in 2015 the Sustainable Development Goals (SDGs) set as one of the targets “by 2020, [to] substantially expand globally the number of scholarships available to developing countries [. . .] for enrolment in higher education [. . .] in developed countries and other developing countries” (United Nations, 2017).
Historically, such assumptions underpinned attempts to “catch up” with the West by various countries as early as the 19th century. A number of Chinese alumni of U.S. universities sent by the Qing Empire as early as 1872 played pivotal roles in institutional reforms of the early twentieth century (Ye, 2002). The Ottoman Constitution of 1876 and the Young Turk Revolution of 1908 that aimed to introduce Western-style parliamentarism were driven largely by cadres who graduated from French or British schools across the empire (Davison, 1961). During the Cold War, study abroad programs constituted a tangible tool in the war of ideas. The advanced industrialized democracies and the communist bloc countries alike maintained varieties of scholarship programs for Third World nationals as well as citizens of countries in the opposing camp (Perna et al., 2014).
Along with increased interest from governments, a growing body of academic work aims to systematically investigate societal impacts of Western training. For instance, evidence suggests that when non-Western, developing countries have Western-educated leaders, this may increase the effectiveness of foreign aid in the long run (Minasyan, 2016), reduce the likelihood of militarized interstate disputes (Barceló, 2020), 1 and increase the chances of economically and politically liberal policies (Nieman & Allamong, 2021). In Ukraine, during and after Euromaidan, a major wave of prodemocratic protests and civil unrest in 2013–2014, Western-educated alumni are believed to have played important roles in launching and consolidating cross-dimensional pro-reform networks (Rabinovych, 2021). In Tunisia, Western-trained parliamentarians may have strengthened legislative support for liberal secular policies (Grewal, 2020).
Certainly, counterarguments to such views abound, as will be discussed later. Given the overarching significance of democracy for many development outcomes—from socioeconomic policies to human rights, military and criminal justice, and general governance (Gerring et al., 2022)—this article asks: What do we know about conditions under which Western training programs foster democracy in the trainee-sending non-Western countries?
Positionality
The views expressed in this study are informed by my background, training, and experiences. Currently an academic and an executive education trainer based at a European institution, I also identify as an immigrant from a postcolonial country. My working-class background, involvement in political activism, and experiences of migration have shaped my perspectives on democracy, social inclusion, and social justice. As a former recipient of a U.S.-government graduate scholarship and other awards from U.S. and Western European donors, I know from personal experience and observation that such opportunities can inspire, empower, and broaden horizons. I have also witnessed and studied the more mundane social, cultural, economic, and political realities that surround these programs, which can intersect in ways that hinder historically marginalized groups.
My experiences as a development aid worker and a student of democratization have made me cautious about embracing either uncomplicated, celebratory arguments or deterministic paradigms concerning sociopolitical change. The former, in my view, too easily attributes excessive agency to individuals while overlooking persistent and subtle oppressive social structures, whereas the latter underestimates the potential for individuals and groups to drive positive social change. My experience as a practitioner of field-based, mixed-methods research and statistical meta-analysis—much like that of many colleagues with similar backgrounds—makes me more likely to notice pitfalls and limitations in social science research that are often downplayed in mainstream academic work. These experiences have also inclined me to value interdisciplinary work, provided it builds firmly on disciplinary insights and is conducted rigorously, rather than as a superficial exercise aligned with passing academic fads.
Core Assumptions
The fundamental observation behind this paper is that making sense of the disparate and conflicting findings from individually forceful studies requires us to move beyond the question of whether Western education fosters democracy in non-Western societies. Instead, we should focus on the more nuanced and consequential question of under what conditions Western training programs support democracy, and under what conditions they do not. This observation is based on three assumptions. For one, as we will see in subsequent sections, different examples can be drawn to show that sometimes Western training may have promoted democracy, sometimes contributed to its antipodes, and perhaps often has remained politically immaterial. As suggested in the background section and as will be further discussed, the landscape of relevant actors, institutions, and interactions between Western and non-Western societies is too vast and too complex to be credibly captured by unidirectional accounts.
Second, even if generally there is an “average effect” of Western training, positive or negative, more or less linear, it is not very helpful for advancing theory and practice because the ultimate effect in real-life settings is likely to depend on varying conditions. As complex interconnectedness among humans makes much of their behavior contextual and conditional, most hypotheses about this behavior are also conditional—contingent on factors such as the resources they have, the institutions they are part of, or the cultural practices dictating their conduct (Brambor et al., 2006; Clark & Golder, 2023; Goodin & Tilly, 2008).
The fact that potential leaders from some countries can pursue Western training, return to their countries of origin, rise to power, and implement liberal reforms but their counterparts in other countries cannot do one or more of these things indicates varying contexts. If the parameters of who applies, who is selected, who returns, who joins civic or political life, who attains leadership, and who pursues political changes are significantly shaped by contextual factors, the political impact attributed to Western training or Western-educated leaders’ agency may be exaggerated, misplaced, and misleading.
Third, the diversity of outcomes is best understood through attending to how they came about. This is probably feasible only if we investigate how an array of contextual factors—relevant initial and background conditions—interact with specific causal pathways in leading to different outcomes (Falleti & Lynch, 2009), in this case, those that link Western training to domestic sociopolitical outcomes. Yet, empirical evidence demonstrating causal relations in existing scholarship is very rare. By identifying contextual factors and their causal effects, conditional theories can transcend optimistic, pessimistic, and skeptical accounts, thereby significantly advancing knowledge and improving practice.
Method
This study combines approaches from critical and integrative review traditions, chosen as the most appropriate knowledge-synthesis methods given the objectives of our research and the disciplinary, theoretical, and methodological diversity in existing scholarship. As a critical review (Grant & Booth, 2009), it aims to transcend description, instead offering a finer degree of analysis and theoretical innovation, exploring contextual variations, and leading to the formulation of new hypotheses and methodological approaches. As an integrative review (Cronin & George, 2023), it intends to expand and diversify the knowledge of this subject by incorporating insights from a variety of different communities of practice engaged in its study and by bridging diverse methodologies. The study was guided throughout by various pieces of methodological advice from Alexander (2020) and S. Wilson and Anagnostopoulos (2021). While adhering to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines, the following style is less formal and more narrative to engage a broader audience, including scholars across social sciences, civil society organizations, donor groups, and policymakers.
Study Selection
The selection of studies was conducted by the author with the help of two pairs of research assistants. We began with two well-defined research questions. First, under what conditions is Western education of non-Western citizens favorable for democracy in these individuals’ home countries? Second, under what conditions is it favorable to nondemocratic alternatives in their home countries?
To ensure a minimum level of quality, we confined our search to academic literature that has undergone quality assessment: peer-reviewed journal articles, books, book chapters in edited volumes (excluding introduction and conclusion), and PhD dissertations. We focused on English-language sources.
We took five steps to capture a broad range of perspectives, uphold citational justice, and incorporate insights from multiple disciplines (for details, see Supplemental Materials, Appendix 1, section 1.7). First, we cast our net wide by conducting a systematic literature search across several electronic databases, including Academic Search Complete, ERIC, ISI Web of Science, and Google Scholar. We aimed to exhaust the search results to ensure that all relevant work was considered and to avoid inadvertently privileging only highly cited or well-known articles, or disproportionately amplifying particularly vocal communities of practice.
Second, the selection of studies was carried out through a dialogic and collaborative process among team members. Third, beyond the baseline quality criteria used in selection, we further reviewed studies based on their theoretical depth and methodological soundness—an approach that acknowledged varying standards of quality, avoided a priori exclusion, and enabled a critical engagement with the literature that was both discerning and inclusive.
Fourth, in addition to the studies identified through initial screening and full-text review, we extended our search through citation chaining. Finally, recognizing that journal publishing, citation practices, and editorial board membership often disadvantage historically marginalized groups, we also deliberately sought out book chapters.
Several criteria for inclusion were used: the studies had to a) contain a focus on Western education, b) provide evidence on non-Western trainees, c) study sociopolitical change in the trainees and/or their home countries, and d) be empirically oriented rather than purely conceptual or theoretical. We formulated keyword combinations that targeted three conceptual domains (for details, see Supplemental Materials, Appendix 1, section 1.5): the explanatory variable (Western education), the dependent or outcome variables (democratic and nondemocratic political change), and the geographical focus (on non-Western countries).
Building on conventional conceptualizations (Cremin, 1976; UNESCO, 2012), we understood education to refer to the systematic effort to transmit, incite, or acquire academic, professional, or social knowledge, skills, beliefs, and attitudes through institutionalized programs of teaching, training, or directed experience. We used a common conceptualization of “Western countries” that includes Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, the United States of America, and Israel (Gift & Krcmaric, 2016). Drawing on Dahl (1989, p. 233), we conceptualize democracy as a political system characterized by seven attributes: universal suffrage and the right to be elected into public office; free, fair, and regular elections; available, enforced, and protected right to free speech; availability of and free access to alternative information that is independent of government control; undeniable rights to form and to join autonomous organizations, including political parties and civic associations; government and party responsiveness to electorate; government and party accountability to election outcomes. We view democratic change as the process of striving for these outcomes through political participation, civic engagement, and public service. Authoritarianism and authoritarian change are the opposites of these. We come back to definitional issues later as an important focus of our analysis.
The selection of studies was systematically structured following the PRISMA 2020 guidelines (Supplemental Materials, Figure S1, online only). The initial search through the databases identified 292 records, from which 104 duplicates were removed, leaving 188 records for further screening. The screening process involved an abstract search that excluded articles based on our criteria, resulting in 65 records proceeding to full-text evaluation. Further searching through citation chaining added 26 studies, and reviewer suggestions contributed three more, culminating in a final selection of 94 studies. Among these, we denoted 39 studies to comprise a “core set” of directly relevant studies (Supplemental Materials, Table S2, online only). The review focuses on them, while also drawing on other included studies.
Two further notes are in order. First, while not every study examined in this analysis is referenced in this article due to space constraints, all sources are documented in Supplemental Materials (Appendix 2, online only) together with examples of studies not included (Appendix 3, online only). Finally, throughout the following analysis, relevant new sources are brought into conversation with the included studies to refine, critique, support, or enrich their findings.
Limitations
While we aimed for a thorough and systematic approach as described, our review has several limitations. First, the reliance on English-language databases and sources could introduce language bias, excluding significant research published in other languages. Secondly, search engines are limited in the literature they retrieve due to indexing biases. In addition, the search terms may have led to our missing articles that would have fit into the purview of the study. We used citation chaining to tackle this. However, this method, while useful for identifying additional sources, could have also led to the inclusion of those that align with the views of the initially selected works (“seeds”). To counter this, we strove to source from a diversity of seeds. Finally, we did not include unpublished works, but some of these may not have been published because of various publication and dissemination biases. Despite these limitations, this review with more than 90 studies, I suggest, provides a representative depiction of the characteristics and trends within the target literature.
Analysis
Building on the thematic-synthesis process (Cronin and George, 2023), I applied a hybrid deductive-inductive thematic analysis approach. I used a priori themes and ideas to guide the analysis, while also inductively deriving new ideas from the reviewed studies. I identified two sets of themes: theoretical themes that reflect implicit or explicit assumptions shaping how research is framed, what is prioritized, and what remains unexamined, and methodological challenges that emerge in the design, implementation, and interpretation of studies (Table 1).
Summary of Themes
The initial themes within these groups were informed by the core assumptions outlined earlier. These themes were operationalized through a systematic historical contextualization of Western trainee trajectories, an element often lacking in prominent studies. In practice, this exercise involved examining the relationship between Western training and purported outcomes by breaking it down into parts (e.g., selection, attendance, attitude change, acquiring networks, return, sociopolitical engagement), assessing theoretical assumptions in each (e.g., to what extent do attitudes actually change?), and unearthing relevant contextual factors (e.g., what factors might foster or impede attitude change?). To understand this relationship further, this exercise also placed it within the wider timeframe of trainee trajectories and the sociopolitical, cultural, and economic contexts surrounding them.
Accordingly, theoretical assumption themes zoom in on key steps within this stylized timeframe, beginning with the exposure to Western education, then follow it chronologically to sociopolitical outcomes while incorporating contextual and conditioning factors from the period prior to, during, and after Western training. Methodological themes start from the most indispensable theory-driven challenges (e.g., Does the study that makes claims about individual-level attitudes measure these attitudes?) to proceed to more sophisticated yet similarly important issues (e.g., [How] does the study examine attitude change over time?).
Initially, primary findings, ideas, and characteristics from each study were extracted. They were classified and abstracted into themes that highlighted common insights and identified potential gaps. Through careful exploration, I considered different ways that these findings, ideas, and characteristics interrelated and connected them back to higher-order themes to further develop the latter. In the final stages, the thematic synthesis focused on integrating themes into a coherent whole. The framework was refined iteratively to ensure it was both comprehensive and efficient, aiming to reveal a clear “big picture” of the topic for informed nonspecialist readers while offering important technical details for specialist stakeholders.
Findings
The Three Core Arguments Distilled
Apart from pursuing immediate and long-term economic motives, Western institutions often provide training to foreigners to exercise “soft power”: an ability to influence the behavior of others through cultural influence, political values, and foreign policies, rather than through coercion or military strength (Nye, 2004). But they also do so to promote proliberal economic reform and democracy in non-Western countries (DAAD, 2016; Dassin, 2013; Dassin et al., 2014; Rice, 2006). One of the five key missions of the German Academic Exchange Service (DAAD) is “to support the process of economic and democratic reform in developing countries and in the transition countries of Middle and Eastern Europe” (DAAD, 2016). Democracy scholars see West-trained individuals as a key channel for international linkages that raise the cost of authoritarianism through higher demand for accountability (Levitsky & Way, 2005). Influential think tanks advocate educational exchanges as a major way for the United States and Europe to foster liberalization in nondemocratic countries (Atkinson, 2010).
Generally, three broad camps offer contrasting empirical perspectives: “optimists,” “pessimists,” and “skeptics.” On one hand, the
Many optimists view the link between Western training and democratization in student sending countries as a phenomenon that holds across countries (Atkinson, 2010; Chankseliani, 2018; Gift & Krcmaric, 2016; Mercier, 2016; Ruby & Gibler, 2010; Spilimbergo, 2009). For some, the conducive factor for democratization is Western education of a variety of individuals (Spilimbergo, 2009). For others, it is the training of state leaders (Gift & Krcmaric, 2016), midranking officials (Freyburg, 2015), or military officers who usually underpin and can subvert autocratic rule and who are believed to become a stabilizing force, particularly in emerging democracies (Atkinson, 2010; Ruby & Gibler, 2010). Socialization into liberal norms and initiation into West-led transnational networks are seen as two key mechanisms through which Western training fosters prodemocratic effects through incentivizing or constraining Western trainees (Atkinson, 2010; Freyburg, 2011; Gift & Krcmaric, 2016).
On the other hand,
Furthermore, Western-educated non-Western citizens, pivotal in market reforms in their home countries—such as the “Chicago boys” in Chile and the “Berkeley mafia” in Indonesia—were elitist technocrats rather than democrats. In fact, they aimed to enhance their state’s autonomy from civil society (Dunning, 2005; Silva, 2010). Recent reports also charge Western-trained cadres from the Persian Gulf to Central Asia with making autocracies resilient through legitimation and know-how (Commins, 2014; Dooley, 2015; Sordi, 2018; The Economist, 2016).
Finally,
While reconciling these divergent views may seem challenging, three key observations can guide us forward. First, despite their various limitations, these perspectives each offer unique strengths. A careful synthesis of these can not only qualify but also complement and refine different findings, thus facilitating a deeper understanding. Second, the stark variation in outcomes suggests that the political effects of Western education are conditional, likely dependent on often unidentified contextual factors. A comprehensive integration of relevant evidence from across the social sciences, including less visible but potentially more nuanced accounts, can help identify and emphasize these variables. Third, although much existing work tends to borrow key concepts or theoretical frameworks flexibly, the presence of alternative theories and evidence both enables and necessitates a push for greater theoretical sophistication in this field. These are the tasks that the following sections will undertake.
Five Theoretical Assumptions: Analysis and Assessment
Assumption 1: Varying Definitions of “Western Education.”
Prima facie, the concept of “Western education” appears unambiguous in the existing literature. To some extent, the concept of “the West” is meaningful, serving as a heuristic that reflects shared historical developments, cultural legacies, and political structures of these countries. Moreover, although the frequent conflation of “the West” and democracy should be analytically separated, this merger can be seen as somewhat justified, given the longer and more sustained experience with democracy in Western countries compared to others.
However, a closer look reveals several issues. First, “Western education” is often defined very broadly, encompassing numerous programs across various Western countries, and its operationalizations frequently rely on Western/non-Western binaries. Some researchers add nuance by breaking it down; for instance, Nieman and Allamong (2021) categorize educational traditions into three groups: Anglo-American, other Western (Humboldtian and Bonapartist), and hierarchical (Latin, post-Soviet, East Asian, and Soviet). Nonetheless, such generalizations risk excessive assumptions of similarity across diverse training destinations, institutions, and programs, juxtaposing them against similarly broad conceptions. On the other hand, defining “Western education” too narrowly, such as training in a specific program, risks making explanations idiographic and not necessarily generalizable to “Western education” as a whole. Finding a middle ground is advisable, but the issue is more profound.
Can we credibly assume, as some seem to do, similar educational approaches across different disciplines 2 so long as they are provided in Western countries? This assumption risks grouping together diverse educational experiences characterized by substantial differences in length, intensity, curricula, and teaching styles among, for example, social sciences, engineering, military studies, and various capacity-building programs. Additionally, such a grouping assumes similarity in West-based educational experiences between programs with substantively different funding schemes. For instance, programs designed and offered by the U.S. or German governments to nurture civic and democratic engagement may differ significantly from those funded by governments of China or Saudi Arabia, which prioritize technical expertise for national development.
More sophisticated approaches involve distinguishing and comparing trainees across various fields, categorized, for example, into economics, law, other social sciences, humanities, and other majors (e.g., Li & Feng, 2018; Nieman & Allamong, 2021). As Nieman and Allamong (2021) reasonably argue, social science and humanities courses may be more likely to expose trainees to classical liberal values. Similarly, more discerning analyses identify potential differences among diverse forms of Western education, such as formal education programs, public diplomacy exchanges, and professional training. Yet, these differences are rarely theorized in depth and even more seldom examined empirically. To be sure, they do share affinities and overlaps (e.g., see Peachey, 2017). However, the risk of inappropriately grouping substantively dissimilar programs and experiences remains high unless we carefully consider their actual contents, as elaborated later.
For John Dewey, true democracy is possible only if the citizens’ education has prepared them to be critical inquirers and conscientious contributors to deliberation (Dewey, 2004). So, we should expect West-based education to be distinctive in this regard. Indeed, it may expose students to curricula that promote democratic principles, give them less politicized, factual knowledge, and introduce them to extracurricular activities that are characterized by open politics (Gift & Krcmaric, 2016). Nieman and Allamong (2021) further argue that, compared to what they group as “other Western” (Humboldtian and Bonapartist), and “hierarchical” (Latin, post-Soviet, East Asian, and Soviet) education models, the “Anglo-American” university model “is characterized by greater autonomy from the state in terms of the course offerings, academic freedom afforded to faculty” (p. 12), more egalitarian instructor–student interactions, diversified instructional methods, and evaluation based on “independent research and critical thinking as opposed to rote memorization” (p. 13).
While plausibly valid, these observations may also be relying on ideal types, saying little on whether the actual contents of the educational program a trainee attends and the quality of their delivery satisfy Dewey’s criteria. It is worth asking: Does the program teach critical thinking and how well, especially since empirically there is no consensus on whether critical thinking can be taught and whether colleges actually teach it (El Soufi & See, 2019; Huber & Kuncel, 2016)? Does this program include a democratic education or civic education component? Is that component strong? 3 Does the training program include community service accompanied by reflection—something believed to be critical in developing commitment to civic engagement (Baxter, 2018)? Conversely, does the program entail elements that nurture authoritarian views, values, and practices, as is assumed by “pessimists”? If we do not account for the actual program contents and quality, we must assume that most West-based programs covered by a study, potentially vast in number, contain these elements and deliver on quality, while the rest—bundled together—do not. If we assume this, for example, by relying on averages, we face ecological fallacy—individuals who studied in any West-based program would be assumed to have had exposure to these elements regardless of their actual experience.
In making such assumptions, it is crucial to distinguish rhetoric from facts and to ensure that illustrative examples are both supplemented and surpassed by systematic empirical analysis of individual-level experiences. Otherwise, we risk discounting facts that may contradict preconceptions or stated ideals. For example, regarding “optimistic” assumptions, critical voices in one leading Western country—the United States—highlight persistent socioeconomic, racial, ethnic, and gender inequalities in access to education (Apugo et al., 2023; Chronicle of Higher Education, 2016; Duncan & Murnane, 2014; Walters, 2001); a heavy influence of political interests and funding on curricula (Carnochan, 1994; Saul, 2018); a disparity between stated democratic ideals and actual authoritarian codes of conduct in schools (Schimmel, 2003); epistemic injustice in doctoral education, where institutional rules marginalize certain knowers and their knowledge claims (Gonzales et al., 2024); increasing managerialism, surveillance, and the undermining of faculty self-governance (Barkawi, 2013; Morales Vazquez & Levin, 2018); and systemic underrepresentation and underperformance of men of color (Cabrera et al., 2022).
Furthermore, conclusions would be more credible if, along with presenting strong reasons to believe in the egalitarian elements of Western training programs, we also consider alternative perspectives—and vice versa. For instance, one might question whether these programs can effectively contribute to genuinely democratic values and democracy if they are inadvertently or deliberately used as tools to assert the inherent epistemological supremacy of modern Western knowledge, fostering academic dependency, and promoting Eurocentric, Orientalist, or neocolonial visions, which are hardly compatible with a broadly conceived democracy (Alatas, 2003; Crossley & Tikly, 2004; Moosavi, 2020a; Tikly, 2004; Watson, 2012). If a program fosters auto-Orientalism in trainees, as evidenced by the acceptance or internalization of Western stereotypes, or by reinforcing narratives that depict their own culture as exotic, backward, or in need of Western leadership, it may promote elitism rather than egalitarianism towards one’s community.
This problem is exacerbated by accumulating evidence of limitations of Western knowledge in serious decolonization work that avoids the “decolonial bandwagon” (Moosavi, 2020b). For one, there is strong evidence that much existing empirical research in behavioral sciences may be skewed because it is often based on samples drawn from Western, educated, industrialized, rich, and Democratic (WEIRD) societies that are not representative of the human species but are outliers (Henrich et al., 2010).
The next concern is timing: Was West-based training received at a “right time” for the trainee to “absorb” it? The impressionable years hypothesis holds that political attitude formation is strongest during adolescence and early adulthood (Alwin & Krosnick, 1991). This may be disputed, but if one of the key assumptions made by the advocates is that socialization into political norms happens because students receive it during their formative years (Gift & Krcmaric, 2016), verifying that trainees were indeed exposed to Western education during a specific time frame becomes important. We also need to assume that Western programs that trainees experience do not change substantively over time, which may or may not be a valid assumption.
Common definitions also seem to conflate educational and social experiences in a Western country. But these are distinct phenomena with conceivably very different effects on one’s socialization. An international student studying at a liberal university in a conservative area in Germany could have very different educational versus social experiences. This relates to the next question. If those approaches that focus on West-based education attribute a causal role to such education per se, would not identical West-rooted online programs and “Westernized” domestic programs in non-Western countries that have similar political effects? Examples include eminent domestic institutions in Japan (which arguably is “Western”), South Korea, and Singapore, increasingly so also in China, India, Brazil, and Mexico, and satellite campuses of major West-based universities in non-Western countries.
For example, institutions such as Central European University (CEU), African School of Economics, and Boğaziçi University have been based in countries that may not be regarded as Western—in Hungary, Benin, and Turkey, respectively. Yet, they have offered programs and experience that are not only “Western” but also on par with many good universities in Western Europe and North America. The case of CEU is curious. Since its inception in 1991, this university has provided quality education that emphasized liberal norms to more than 14 thousand alumni, particularly from central and Eastern Europe and the former Soviet Union, many of whom have joined politics in their countries of origin and rose to prominence, often on a platform of liberal reforms. CEU was forced out of Hungary due to pressure from Hungary’s Prime Minister Viktor Orban, who, ironically, spent a semester studying at Oxford University on funding from CEU’s founder and major advocate of open societies, billionaire George Soros. It would be telling to examine whether—and in what ways—the student experience changed following CEU’s move to Vienna. Did its educational approach become “more Western” as a result of its relocation to Austria? 4
Assumption 2: Questioning the Expectation of Attitude Change
A key mechanism through which the political effect of Western education is believed to work is future leaders’ socialization into particular political views, values, and practices. While some studies see university as key (Gift & Krcmaric, 2016; Nieman & Allamong, 2021), others argue that these changes happen only if the training program involves practical experience and direct contact with civic and political institutions, but not if more diffuse contacts, such as international education, are involved (Atkinson, 2010; Freyburg, 2015). However, few studies examine changes in views and attitudes empirically and on the individual level, which seems the most appropriate level to study them. Refreshing yet partial exceptions, such as Freyburg (2015) and Han and Zweig (2010), often study attitude differences among groups in a cross-sectional setup, but this does not measure attitude change for individuals. If attitude changes are important (but see Teorell & Hadenius, 2006), verifying them is critical. Here, too, reconceptualization and a deeper empirical look are essential.
First, if effective changes in political institutions are profound social transformations requiring changes in culture (e.g., values, norms) and/or social structure (e.g., power, class structure) as extensive social science literature suggests (for illuminating syntheses, see Portes, 2006, 2010), political socialization during Western education needs to be profound and overcome higher potential hindrances than currently believed. First, prestudy background, values, and experiences can condition, if not block, attitude change. Differences in these can entail differences in study outcomes. For example, previous attitudes, prior experience, and family environments impact adaptation experiences and subsequently exchange outcomes (McKay et al., 2022; Sijapati & Hermann, 2012). While some trainees become “integrators,” others who have moderate cultural flexibility become “adapters,” and those who are culturally less flexible become “isolators” (Murillo, 2021).
Durkin (2008, p. 15) suggests that, in the case of critical thinking skills, the majority of trainees opt for a “middle way” that “synergizes their own cultural approach to critical thinking with those aspects of Western-style critical thinking and debate that are culturally acceptable to them.” Local agents often modify foreign norms to align them with their existing cognitive frameworks and identities (Acharya, 2004). Migration literature also highlights the resilience of premigration values, visible, for example, in the tendency of diasporas to reproduce home-country political divisions (Ahmadov & Sasse, 2015; Lafleur & Sánchez-Domínguez, 2014).
Second, many scholars and practitioners seem to assume that non-Western participants see the values, norms, and institutions of Western countries as superior to those in their home countries. While probably true in some cases, there is no consensus on whether this is a pattern. For example, Peachey (2017) shows that alumni of the Future Leaders Exchange (FLEX) program, a prestigious U.S. scholarship scheme for young citizens of Eurasian countries, do not have an idealized view of the United States before and especially after the completion of the program. Instead, they have “nuanced and critical perspective on the U.S. government” and “a less favorable attitude toward the United States government than funders would desire” (2017, p. 73). Furthermore, FLEX alumni seldom viewed the study outcomes the way the funders envisaged, such as in terms of socialization into norms and building transnational networks (Peachey, 2017, p. xiii). In another example, substantial frustration over poor working conditions at host universities significantly harmed the post-Fulbright perceptions of African Fulbright scholars (Sunal & Sunal, 1991).
When social experience is concerned and, for instance, positive experience is assumed, such assumptions risk ignoring problems with Western democracies that may not escape visitor’s attention, such as the dominance of organized interests (Hacker & Pierson, 2014), failure to redistribute income that props public support for the Right parties (Flaherty & Rogowski, 2021), and political exclusion of labor migrants that may be an essential part of the embedded liberal compromise (Goodman & Pepinsky, 2021). Trainees can see that in some Western societies, to paraphrase Schattschneider, “the heavenly chorus” of ostensibly liberal democracy “sings with a strong upper-class accent” (Schattschneider, 1960, p. 35). While exposure to Western political institutions can win over some students, it can also disillusion others (I. Wilson, 2017).
To have a political effect on the sending country, Western training programs should win over particular subgroups, such as potential future elites (I. Wilson, 2016). However, when “democratic transfer” for such subgroups is assumed, one could question the extent to which many such individuals, such as scions of elites, “soak and poke” sufficiently, if at all, in the venues, practices, and institutions where civic and democratic learning in Western countries may take place most, be it local institutions, campus, or classroom.
In some cases, studying in Western societies and being exposed to their institutions may also lead to a rejection of liberal values—as in the case of many Western jihadists (Gambetta & Hertog, 2009; Kapur, 2014). While some Chinese students surveyed in one study developed favorable attitudes toward their Western hosts, others developed hostility (C. Li, 2010). Hand and Zweig (2010, p. 300) found that Chinese “returnees from Canada are more hawkish than returnees from Japan.” Institutional and cultural barriers can preclude migrants, including students, from socializing into political practices and norms in their host countries (Careja & Emmenegger, 2012; Sijapati & Hermann, 2012; Vogel & von Ossietzky, 2005), and their views can be affected by individual experiences of freedoms and constraints regardless of destination (Rother, 2009).
Furthermore, in a so-called “ceiling effect,” the training might fall short of producing the envisaged impact because the participants already have the attitudes intended by sponsors. Such a phenomenon has been observed in measuring whether the Erasmus program makes its participants feel more European (Kuhn, 2012; I. Wilson, 2011). This is particularly important to study given reverse causality concerns (I. Wilson, 2016): Western education programs might be selecting individuals who already possess attitudes that the program intends to equip them with. This may not be true across the board. Many student migrants can be driven by a desire to get high-quality training and be indifferent toward what they see as politics. Scions of nondemocratic country elites who have higher access to Western-based training may be unlikely to hold liberal views (Nieman & Allamong, 2021). However, explicit theorizing and individual-level empirical investigation of this issue are crucial for supporting causal claims, including when self-selection may not necessarily be based on overt ideological position but rather on attitude plasticity.
Finally, the durability of attitude change needs consideration. When attitudes indeed shift while abroad, they need to endure to affect politics back home in the direction of that shift. Yet, as early as the 1950s, evaluators of exchange programs found that when attitudes indeed seem to have changed, this may not last (J. Watson & Lippitt, 1958). Further, trainees’ evaluation of their Western host country politics partially depends on their evaluation of their home-country politics, but the latter can be unstable, thus resulting in reevaluation of the host country. For example, home-country assessment may follow a U-shape curve: one study finds that while initially the trainees’ evaluation of home-country conditions and its future prospects deteriorates, later it tends to rebound (Li & Feng, 2018).
Assumption 3: Dominance of Western Cultural Capital and Networks
While attitude and value changes may be important, they may not be sufficient for producing political outcomes, particularly if trainees represent underprivileged groups with few chances of occupying leadership positions upon return. While some studies of the political impact of Western education do not go beyond values, others do. For example, those arguing for the prodemocratic effect suggest that it partly works through leaders’ initiation into Western-led, transnational networks that simultaneously enable Western support for democratization and act as a leverage on these leaders (Atkinson, 2010; Freyburg, 2011; Gift & Krcmaric, 2016). While such hypotheses rightly underscore the importance of social networks as potential causal pathways, these explanations need nuancing against several considerations.
First, Western support for democracy in non-Western countries is neither a given nor homogeneous. While foreign policies of various Western democracies have had democratizing intentions and effects in some places, they also have included autocratic regime support and antidemocratic effects elsewhere (Carothers & de Gramont, 2013; Monga, 1997; Schmitz, 2006). On their end, non-Western kleptocratic elites, such as those from Eurasia and Africa, utilize transnational linkages of “uncivil society” that are often sourced from Western countries to pursue strategies of reputation laundering (Cooley et al., 2023).
Second, the assumption that emphasizes transnational linkages—democratic or antidemocratic alike—risks underplaying the role of domestic networks, which are likely to be more critical for upward political mobility back home. For example, for galvanizing prodemocratic action, Western education would need to affect home-country social structures in a way that undermines the traditional intergenerational transmission of resources and power, which inhibits the emergence of new elites. For this, it should increase participants’ cultural and social capital in ways that increase their influence back home. Whether this is the case is an open question.
Cultural capital includes qualifications, accumulated knowledge, skill repertoires, tastes, and material belongings (Bourdieu, 1986, p. 82). Social capital, in Bourdieusian sense, refers to “the aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition” (Bourdieu, 1986, p. 86). Cultural and social capital accumulated during and due to Western education should be appropriate—and not counterproductive—for gaining upward mobility, given the requirements of the home country’s social and political structures. Otherwise, Western-educated individuals are not likely to make it through domestic elite selection structures to become political leaders in the first place. On a basic level, cultural divergences between alumni and their origin-country communities, manifested, for example, in stereotyped alumni mannerisms, can limit the impact that these alumni can make in their home community (E. C Wilson & Bonilla, 1955).
Cultural and social capital gained abroad should also be sufficiently high to boost the recipients’ upward mobility in the home country, given the opportunity costs of absence there and the home-country competition they face. When Western trainees leave their home country, they are likely to temporarily forgo opportunities to tap into domestic social networks whose support would be conducive for their later rise into positions of influence. Opportunity costs can be high as competitors for the same positions can take over.
The record of Western programs in terms of their effects on the accumulation of cultural and social capital is mixed at best and remains murky due to a lack of data. Since only individuals with good home-country education usually match the criteria to be selected into Western programs and they are not likely to come from underprivileged backgrounds, these programs can help reproduce social inequalities (Wächter, 2012) or, at best, “hereditary meritocracies” (The Economist, 2015). Evidence from Europe and Latin America suggests that individuals from families with low educational attainment; part-time students; women; and Indigenous, Afro-descendant, or other marginalized groups may not meet the requirements or might be less likely to apply to study abroad even if they do (Dassin, 2013; Netz, 2013). To have a genuine democratizing effect, Western programs would need to augment the cultural and social capital of alternative elites who may have higher incentives to contribute to democratization than current elites. Programs like the Ford Foundation International Fellowships Program aim to empower the underprivileged with the goal of tackling social inequality, but report mixed results (Dassin et al., 2014).
Third, to build Western-based transnational political and professional linkages that may later support their civic and political engagement, trainees need to develop them while abroad and through their (Western) education. However, since Bochner et al. (1977) and Furnham and Alibhai (1985), it has been shown that foreign students strongly prefer conational networks of friendship followed by multinational second and host national third. While the function of the first type of network is cultural affirmation, the others mostly serve to facilitate recreational and academic motivations. Subsequent evidence varies considerably, rather than confirming the dominance of transnational linkages with Western hosts. Finally, alumni also lose contact fairly quickly and, even if professional or political linkages were built, finding them years later has proven difficult (I. Wilson, 2017).
Assumption 4: Sociopolitical Engagement Back Home?
A natural next step for Western education to have a political effect is that the recipients, in fact, contribute to civic and political processes back home. Several gaps in existing knowledge defy assuming this away. First, although Western-trained individuals may not need to return to their home countries to affect home-country politics as they can influence political preferences and outcomes from abroad (Burgess, 2012), the return brings them much closer to making an impact (A. C. Campbell, 2020). Yet, while existing studies on the political effects of Western education are usually based on evidence from returnees, such returns are not a given (Tessema et al., 2012).
While some programs have a strong conditionality element that requires participants to return home and contribute to sociopolitical processes, public service, or business, many others do not (A. C. Campbell, 2018). Networks of “brain circulation” between Western and non-Western countries are rare. According to UNESCO, “Out of 22 Latin American and Caribbean countries, 8 maintain formal brain circulation networks” (2021, p. 343). Theoretically, if Western trainees are highly motivated social change makers with knowledge and skills to contribute to democratization or improved governance, their “brain drain” may weaken pressures for democratization or hinder democratic consolidation back home. Furthermore, if the return depends on factors that also predict the likelihood of political change, we might inaccurately attribute political effect solely to returnees rather than to the underlying variables that enabled or fostered their return and political participation.
Third, if Western trainees return home, to have an impact, they need to engage in civic and political processes. The link here is also not automatic. Also, education may shape attitudes, but not necessarily subsequent behavior: Attitude change can fail to translate into expected action (Cantoni et al., 2017). The case of Aung San Suu Kyi, once Myanmar’s iconic British-educated leader, is illustrative, if not puzzling. Hailed in many countries as a prodemocratic hero, her reputation was shattered when she failed to act to respond to the massacres of Rohingya people in her country and refused to acknowledge that Myanmar’s military committed these atrocities.
Furthermore, higher educational attainment, which Western-based institutions can significantly add to, may sometimes promote and sometimes hinder civic and political engagement. While earlier theories argued that higher levels of education are more conducive to civic and political participation, this argument has since been strongly questioned (D. E. Campbell, 2009; Kam & Palmer, 2008). In addition, the effect of scholarship conditionality on trainees’ interest and involvement in social change is also not straightforward: It can promote or limit them depending on overlap or tensions between program goals and trainees’ personal plans (A. C. Campbell, 2018).
Finally, when Western program alumni get engaged in civil society and politics, we should not assume that it is prodemocratic. While this is true in some cases, Western alumni can also contribute to legitimation and consolidation of the home-country’s nondemocratic institutions (Kapur, 2014; Silva, 2010; Sordi, 2018). For example, between 1994 and 2018, Kazakhstan funded more than 12,000 students through its “Bolashak” mobility program, with many studying in Western democracies. However, while successful in terms of professionalization of cadres and career mobility for its recipients, the program may have also contributed to authoritarian stability, largely through cooptation and legitimation (Sordi, 2018).
Assumption 5: Homogenized “Home Country”?
While some studies indicate home-country barriers for Western trainees’ civic and political engagement back home, others seem to attribute a notable role to individual agency. Yet, despite the considerable influence of cross-border flows of ideas, knowledge, and people, nation-states remain key players in structuring these flows through various policies, such as migration control and development cooperation (Faist, 2008). The assumption of a smooth global-to-local flow of norms inherent in some studies appears unwarranted (Zwingel, 2012). Furthermore, rhetorical adaptation is sometimes employed to resist the international flow of norms (Dixon, 2017).
Largely underspecified and underexplored in the existing literature on political impacts of Western education, varying home-country conditions can explain why, to begin with, in some countries, potential prodemocratic leaders can leave their homeland for Western destinations and are able to return, rise to power, and implement reforms, but in other countries, their counterparts cannot or do not do the same. Therefore, home-country factors should be incorporated into theoretical and empirical models in ways that account for social structures and filters. To be sure, some cross-national studies (e.g., Nieman & Allamong, 2021) statistically examine whether home-country factors like previous liberal reform make it easier for liberal-minded politicians to enter office by using previous reform as a predictor of the subsequent leader’s education. Yet, such tests remain substantively undeveloped to do justice to the variegated nature and sequentially complex influence of home-country political-economic context on pre- and post-training incentives, behavior, and trajectories. As the extensive review of scholarship evaluations argues, for progress in understanding, the focus should be on explicating complex questions on how the outcomes of Western training come to be (Mawer, 2018).
Drawing on migration literature on push and pull factors, three groups of home-country conditions may be particularly important to investigate. The first relates to the quality of governance, particularly the political regime. Specifically, less democratic and politically open polities can encourage departure, discourage return, limit political participation, undermine prodemocratic alumni, and curb their contribution to democratization processes. Evidence suggests that autocratic leaders can strategically manipulate emigration policy by liberalizing it for economic motives or by restricting it if they fear that their emigrant citizens in democratic countries can destabilize their political regimes (Miller & Peters, 2018).
Conversely, a higher degree of political openness, for example, can stimulate alumni networks, which individual pro-reform alumni find important in enabling and facilitating their efforts to “give back” through social and economic change, and without which many do not engage in civic or political processes (A. C. Campbell, 2016). In addition, the extent to which political and economic fields are meritocratic (Bonilla & Kwak, 2015; Dassin, 2013), which can be related to constraints on ruling elites and on government corruption levels, can define upward mobility and alumni’s rise to positions of influence. Existing reforms, government effectiveness, and work conditions can shape alumni perceptions of the attractiveness of pursuing government employment (A. C. Campbell et al., 2021) or contributing to socioeconomic development (A. C. Campbell, 2019; Z. Wang et al., 2024).
The second group of factors is home-country economic conditions. Economic diversity and openness can increase job opportunities, encourage return, and strengthen prodemocratic alumni’s capacity to mobilize. Economic development and the absorptive capacity of the economy affect the availability of jobs and opportunities for mobilizing through the workplace (Wächter, 2012). For example, developing countries that suffer from deindustrialization have fewer job opportunities and fewer linkages across sectors that have been shown to be conducive to political mobilization (Dunning, 2005). This can discourage return, encourage brain drain, and limit alumni capacity to mobilize actors and affect change. Case studies of Western alumni from diverse countries, such as Eritrea, Kazakhstan, Ghana, Guatemala, Nigeria, and China, provide empirical support for these ideas (Bonilla & Kwak, 2015; A. C. Campbell et al., 2021; Fu & Zhao, 2017; Jonbekova et al., 2022; Tessema et al., 2012).
Finally, the strong affinity of elites and publics with Western countries can enhance upward mobility for Western-educated alumni. Improving cultural understanding through exchange programs is limited if the relations between the home and host countries are strained, and the alumni of such programs are marginalized (Scott-Smith, 2008). Furthermore, when alumni exhibit stereotypical behaviors of their Western hosts that reflect cultural differences between them and their home-country communities, it limits the influence they can have back home (E. C. Wilson & Bonilla, 1955).
Methodological Challenges and Strategies for Addressing Them
What methodological challenges arise in the design, implementation, and interpretation of studies on the political effects of Western educational programs in non-Western countries that hinder the ability to draw robust conclusions? What strategies can be employed to address these challenges?
Challenge 1: Systematically Zooming in on Individuals
If key mechanisms through which Western education is believed to have political effects are changes in Western trainees’ attitudes and their initiation into transnational networks, assembling and analyzing individual-level evidence is critical to shed light on these pathways. This is a major methodological challenge in moving forward. Existing scholarship shows a mixed record here.
One complication comes from a common practice of using aggregate country-level variables as proxies for individual behavior. For example, gross numbers of West-bound students per country are used as predictors of political regime, liberal reforms, and similar political outcomes (see Atkinson, 2010; Spilimbergo, 2009). While useful for detecting associations and attaining generality, such approaches often leave causality underexplored even when conventional statistical techniques are enhanced with refinements (Seawright, 2010). Nor will they discard alternative hypotheses convincingly.
Second, some studies use the education of state leaders as a predictor of country-level political outcomes. This approach contributes to a welcome revival of interest in the political, economic, and social effects of leaders’ biographical characteristics. However, it also raises two concerns. First, focusing predominantly on rulers overlooks various layers of selection previously discussed, potentially misattributing causal significance to what might be a secondary factor. Second, such approaches often do not empirically address alternative arguments that Western-trained midranking officials, military officers, and individual citizens play critical roles in political change (Freyburg, 2011; Ruby & Gibler, 2010; Spilimbergo, 2009).
Third, while small sample sizes are sometimes unavoidable or justified for in-depth analysis, a more significant issue in studies focusing on individual subjects is the selection of participants. These often rely on convenience sampling, with insufficient detail provided about the sampling process to clearly understand inference limitations. Furthermore, larger studies typically concentrate on a few programs or destinations without comparing different groups. For example, an excellent study by Peachey (2017) examines 1,135 finalists from one program, and Dzotsenidze (2021) compares 50 U.S.-educated Georgian alumni with 99 locally educated students. Although valuable insights emerge from such research, the lack of systematic inter-group comparisons weakens causal claims and hampers the accumulation of systematic knowledge.
Sampling will remain a major issue given difficulties in sampling people on the move, response rates are likely to remain modest, and patterns of nonresponse are difficult to investigate given the reliance on intermediaries, such as training organizations and institutions, in approaching prospective respondents (I. Wilson, 2016). However, tangible steps toward increasing heterogeneity in sampling and subsample sizes sufficient for meaningful comparative analyses can facilitate stronger conclusions. For one, we can distinguish and sample Western trainees in different groups, such as students in different levels of education (e.g., high school, undergraduate, graduate, postgraduate), scholars, and professionals and practitioners. Among other things, this can help understand whether and how age and/or different program type condition attitude or behavioral change.
Furthermore, since exposure to political socialization can differ by funder, as they often have different goals, setups, and practices, distinguishing funding sources is worthwhile. For example, four groups can be differentiated: those on publicly funded Western exchange or training programs (e.g., via British Council, DAAD, various EU capacity-building programs, Fulbright), those funded by origin-country governments, those funded by private organizations or universities, and the self-funded. Since many programs conduct orientation meetings, these can serve as efficient venues for intercept point survey, which can be effective in sampling and surveying migrants (McKenzie & Mistiaen, 2009).
Furthermore, systematically relating sample properties to the population of interest is essential. Country-level population and subpopulation sizes of internationally mobile trainees can be approximated from UNESCO statistics on internationally mobile students as well as from population and education variables in world development indicators and specific national statistics. Program-level statistics are often available from donors. Adopting a mix of sampling techniques, such as heterogeneity and quota sampling, with the population data in mind, may help increase—or at least clarify—the inferential leverage of a study.
Finally, the literature may benefit from greater mindfulness regarding which perspectives are included or excluded. There is a notable geographic imbalance in both the focus of the studies and the origins of the authors involved. While there is a significant variation in the students’ home countries and regions studied, the United States emerges as the predominant host country, followed by some focus on the EU, often overshadowing other Western destinations, which are less frequently studied (Supplemental Materials, Table S2, online only). This distribution suggests a potential geographic bias that needs more attention. Moreover, author affiliations predominantly involve North American, primarily U.S.-based, and Western European researchers, with comparatively fewer contributions from other regions. Factors such as higher economic development and unequal resources partly explain this skew, yet the latter may nonetheless limit the range of insights that can be drawn from this literature. This highlights the need to incorporate diverse viewpoints from underrepresented voices, countries, and regions.
Challenge 2: Contrasting With “Control” Groups
To approach causal claims, it is important to compare Western-trained individuals to “control groups.” Such “counterfactuals” are largely missing in existing work. Partial exceptions—such as Dzotsenidze (2021), who compares the civic engagement of Georgian alumni of two U.S. programs to that of a group of master’s students in Georgia, and Han and Zweig (2010), who compare Chinese returnees from Japan and Canada—are limited in their coverage of programs and control groups. Comparative designs that provide some “counterfactual” evidence, with the preferred group for comparison being nonselected finalists, are more prevalent in program evaluations (Mawer, 2014). Here, the resource drain involved in identifying appropriate groups for comparison, potential lack of cooperation from nonselected finalists (if they can be identified), and the assumption that time-varying factors are constant across the selected and nonselected groups are key concerns (Mawer, 2014).
A systematic comparison with specific “control groups” could help mitigate several issues. The first group is individuals heading to non-Western destinations for training. Their incorporation can increase confidence that the effect, if any, attributed to Western education is not due to foreign education in general. The second are “close contacts” in the home country—individuals from similar backgrounds as West-bound individuals, but who do not have foreign education and who remain in their home country for the foreseeable future. Including this group may help have a baseline for comparison. The third group is members of the general population who “match” West-bound individuals in their demographic and socioeconomic profiles. Their inclusion would allow us to see whether changes in West-bound trainees are due to shifts in the origin country that affect larger segments of the society. The fourth are individuals in home-country Westernized institutions and online Western courses. Their inclusion would tell us whether physically studying in Western countries is imperative.
While sampling these groups is difficult, creative solutions that maximize heterogeneity of sample composition and reflect local contexts may help. For example, West-bound individuals can be asked to provide their “close contacts”—friends or colleagues who are matched on various characteristics (e.g., age, education level, place of origin within the home country, gender, occupation), who do not have foreign education, and who the respondents believe are likely to remain in their home country in the meantime. To draw samples from the population and Westernized institutions, multistage sampling may prove useful. For example, within a country, researchers can identify a selection of settlements and institutions from which many Western trainees come and sample individuals from those settings. To sample individuals in online training courses, one could create sampling frames from online communities of potentially eligible people and sample from within these frames. If respondents in different groups are recruited through different sampling techniques, developing weights for respondents in different groups and adjusting the analysis for these differential weights could reduce resulting biases.
Challenge 3: Embracing Time and Temporality
Existing studies are predominantly cross-sectional, ex-post studies. Some face challenges in reconstructing baseline data from retrospective measures, and many lack baselines altogether (Mawer, 2014, 2018). Yet, without a longitudinal design, we can hardly measure changes in values, rule out selection effects, and investigate continuity or change.
Multiple-wave studies of Western trainees and control groups can tackle such issues. The first wave would set the benchmark. The second wave would ensure that subjects are indeed exposed to what we think they are exposed and that a change or stability in respondents’ views or practices is measured and not simply inferred (e.g., based on their location). Subsequent waves would measure the stability of changes, if any.
While panel studies would be ideal, the key obstacle is the sheer difficulty of designing and conducting such studies, particularly with people on the move. It can be hard to retain them, resulting in high panel attrition. However, cohort studies can be feasible. Furthermore, drawing on Deaton (1985) and Deng et al. (2013), one could combine a panel with a cohort study with repeated cross-sectional draws of “refreshment samples” in subsequent waves. This may tackle issues arising from panel attrition. It can also reduce the risk of “panel conditioning”: when participation in one wave affects participants’ answers in subsequent waves (Halpern-Manners et al., 2014).
Challenge 4: Comparing Home-Country Contexts
While well-designed, large-N, cross-national studies can help incorporate home-country context into analysis, wide geographic coverage can significantly reduce internal validity and seriously hinder causal inferences. Focusing on a set of countries can tackle these issues. Indeed, some studies of the sociopolitical impact of Western programs adopt in-depth, small-N qualitative comparative designs (A. C. Campbell, 2016, 2017; A. C. Campbell et al., 2021). However, existing work lacks a strong record on maximizing internal validity and on providing a theoretically articulate and methodologically sound framework to assess the role of varying home-country contexts.
Comparisons of cases chosen from the universe of non-Western countries through careful most-similar systems design (MSSD) are probably the most promising option. Such comparative cross-country analysis can reduce conceptual stretching, enhance the likelihood of correctly identifying causal factors, and allow tracing and assessing causal processes at closer range (Brady & Collier, 2004). To reap benefits from such designs, this field should be sensitive to three issues. First, the chosen (country) cases should share those historical, social, economic, and institutional similarities that can be confounders for the particular outcome under study. This ensures a high degree of similarity in country-level context to discern whether specific hypothesized contextual differences among these countries led to differences in the extent to which Western trainees could impact political outcomes back home.
Second, the selected cases should be equally likely to have been open to and receive Western educational opportunities from the outset. Variations in openness to Western programs could complicate causal inference, as it might suggest that differences in political outcomes are driven by domestic factors rather than Western education.
Finally, unlike large-N, cross-national studies where geographic expertise may not be pivotal though is invariably useful, most-similar systems design should reflect such expertise to significantly enhance study quality and feasibility. Once successfully conducted in one geographic setting, such studies can be replicated in others, ideally in a multi-investigator, cross-regional comparisons that bring together area experts, hence maintaining internal validity while also increasing generalizability.
Challenge 5: Upgrading Methods and Measurements
Finally, current scholarship displays substantial room for improving methods and measurements. On the positive side, there is a visible diversity of approaches and methods. For example, many alumni tracer and evaluation studies adopt mixed-methods approaches, and few rely exclusively on quantitative or qualitative methods (Mawer, 2014). Mixed-methods approaches could strengthen academic studies as well, since the variegated impacts of educational programs are more likely to be captured by multiple tools and triangulation than single-method approaches. Furthermore, significant overlaps in topics among most evaluations—such as identification of sociodemographic profiles, scholarship experience, home-country return, and post-study societal contributions (Mawer, 2014)—promotes cross-fertilization and accumulation of knowledge through exploring meta-level patterns. Furthermore, newer studies show increasing theoretical and methodological sophistication.
However, certain approaches are prevalent, and this creates serious limitations. Apart from cross-national, large-N studies with aggregate data, interviews and surveys are dominant approaches. But conventional surveys have potentially critical difficulties—for example, in tackling social desirability bias: participants tend to give answers that they think the interviewer wants to hear. Most studies are also ex-post cross-sectional, done after the experience and without a longitudinal component and baseline. If such a study is survey-based, for example, interviewees may overestimate the degree of change in their attitudes over time, and we would not know how to check this (I. Wilson, 2016).
Although their feasibility and ethical aspects are major challenges in this field, multisite randomized controlled trials with subsequent meta-analyses that synthesize their results could alleviate outstanding questions regarding causality. However, more accessible and feasible approaches can help too. For one, ex-ante or prospective longitudinal surveys involving different control groups would be a significant step forward.
Second, findings from participant surveys—particularly those relying on self-assessment—could be significantly strengthened if complemented and triangulated by evidence from more systematic surveys or interviews of experts and stakeholders than currently available. In particular, the latter work can be improved through stronger theoretical frameworks and more systematic sampling. Theory-wise, such interviews can be designed to cast critical light on preferences and ability of different citizens to study in Western countries, Western-educated alumni’s pre- and post-study political attitudes and resources, their actual political participation, their actual political impact, and on what this impact depends. Sampling-wise, different categories of experts and stakeholders—such as funders, international development workers, journalists, researchers, and government officials—should be distinguished rather than lumped together to capitalize on their strengths in providing insights into specific domains.
Third, qualitative historical analysis could help overcome several limitations if it offers a systematic macrolevel view in a detailed, theory-guided, and comparative design. Why? Even when repeated for several waves, surveys may not provide evidence on whether the trainees contribute to political processes in the medium or long term, what home-country conditions affect their behavior, and whether they are able to make an impact. They can also be weak in telling us whether the values and resources developed in Western countries were, in fact, instrumental for the alumni’s contribution and their rise to positions of influence. Also, the impact of educational programs may be indirect and take a long time to mature.
While some of the studies reviewed offer a historical perspective, others are situated within specific, time-bound contexts (see Table S2, online only). Furthermore, although certain studies examine relatively long time spans—for instance, a quantitative study by Atkinson (2010) covers data from 1980 to 2006—they do not necessarily trace social processes and transformations over time to understand how past contexts shape present attitudes, behavior, and institutions.
Attention to time is important in another way. Temporal discontinuity and nonlinearity in international students’ experiences (Shahjahan et al., 2025) 5 are likely to extend well into their post-study trajectories. As previously noted, alumni often lose contact quickly, the durability of attitude change gained abroad is uncertain, and various factors can derail initial aspirations for civic or political engagement. Moreover, the accumulation and then erosion of social and cultural capital across contexts—gained at home, weakened abroad, and then vice versa—further illustrate these temporal complexities. Nonlinearity can be seen within individual trajectories, such as in the cases of Aung San Suu Kyi and Mikheil Saakashvili, whose early democratic commitments shifted significantly over time. Furthermore, temporal discontinuity can amplify the role of privilege in shaping selection across the post-study trajectory: those who return, engage politically, rise to leadership, and possess the agency to influence change are often those who had the resources to be “filtered in,” while others—due to structural or contextual constraints—are “filtered out” at various stages along the way.
Among other things, qualitative historical analysis can help discern how alumni self-assessment matches their actual behavior. Such an analysis could draw on documentary evidence, secondary sources, statistical records, recollections, and interviews. Evidence from multiple sources would be combined in process tracing of causal mechanisms and processes, including reconstructing major events that shifted political outcomes, to understand the role of Western trainees in comparison to other groups.
Finally, instruments across studies differ greatly in, for example, how they measure political participation, educational experience, attitudes, resources, and impact. While some make use of major surveys to enable comparison (e.g., Peachey, 2017), many develop their own measures. This is often understandable because such study-level measures aim to reflect the context and the aims of the study. But some cross-fertilization of measures can help advance the accumulation of knowledge and inform policy better. In surveys, for example, a clear advantage of borrowing from existing established measures—whether from the Afrobarometer, Arab Barometer, Latinobarómetro, the European Social Survey, or the World Values Survey—is the potential improvement in reliability as their questions would have been tested multiple times. Such borrowing can also make comparisons easier as well as cut costs—for example, with the general population or specific groups. Particular attention should be paid to avoid motivated response biases—something that is also largely missing in the literature—for example, by considering using less explicit measures or even implicit association tests.
Conclusion
This review identified major themes and charted new frontiers for understanding the impacts of Western training programs on democracy in non-Western societies (Table 2). Based on this analysis, one “conditional agency” framework suggested by this study involves different groups of non-Western country citizens encountering varying opportunity and constraint structures to receive Western training, return to their home country, participate in political or civic life, rise to positions of influence, and pursue political agendas. Similarly, one “prodemocratic summary” of this article may be that to help democratize political institutions in non-Western countries through Western trainees, Western educational experience should cause a durable prodemocratic change in the trainees’ political values, equip them with social and cultural capital that is conducive—not counterproductive—for their upward mobility back home, incentivize alumni to contribute to prodemocratic process rather being coopted into social structures that legitimize and stabilize nondemocratic institutions, and overcome obstacles in home-country context that probably keep nondemocratic governance in place.
Findings Emerging From the Thematic Analysis
If contextual factors are adequately considered, Western education programs can have significant potential to promote political freedom and political equality in developing countries. Then Western and non-Western actors committed to democracy should carefully and thoughtfully explore expanding, restructuring, or modifying such programs accordingly. Such understanding can also better inform policies in several critical related areas identified as priorities by Sustainable Development Goals, various governments, the European Union, and multilateral organizations. These include improving access to quality education, combating social inequalities, strengthening accountability institutions, and enhancing constructive engagement between Western and non-Western, Global North and global South countries. Otherwise, there is a high risk of perpetuating political inequalities and social exclusion in the home countries of Western trainees, as well as reinforcing the narratives and practices of anti-Western movements. Organizations such as the World Bank, the largest global education funder that shapes reform through funding, knowledge, and policy influence (Edwards et al., 2024), should take these risks seriously.
While historical in orientation, this review synthesizes studies anchored in the century spanning the 1910s to the 2010s. Although its temporal scope covers multiple decades—and indeed a full century—the conditions it identifies and analyzes may be shaped by this specific historical framing. As factors such as the expansion of virtual and hybrid learning models, increased migration and forced displacement due to climate change, the resurgence of realpolitik, rising geopolitical tensions, and changes to visa regulations and post-study work opportunities continue to evolve, they are likely to reshape patterns of student mobility, graduates’ career trajectories, and democratic development. Future research will need to account for these and other emerging conditions. 6
Supplemental Material
sj-docx-1-rer-10.3102_00346543261426599 – Supplemental material for Western Promises? A Critical Review of the (Anti-)Democratic Effects of Western Education in Non-Western Countries
Supplemental material, sj-docx-1-rer-10.3102_00346543261426599 for Western Promises? A Critical Review of the (Anti-)Democratic Effects of Western Education in Non-Western Countries by Anar K. Ahmadov in Review of Educational Research
Footnotes
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: I would like to gratefully acknowledge the Dutch Research Council (NWO) Vidi Talent Programme Grant VI.Vidi.191.194. The ideas in this article have benefited from discussions with Bob Keohane, James Hughes, Farid Guliyev, and Judi Mesman, as well as conversations with fellows and colleagues at the Netherlands Institute for Advanced Study (NIAS) during my fellowship there in 2020–2021. Finally, I thank Aray Gaipova, Vadim Lacroix, Kato Orjonikidze, Douwe Roest, and Elvin Yusifli for their excellent research assistance.
Notes
Author
ANAR K. AHMADOV is associate professor at Leiden University, Anna van Buerenplein 301, 2595 DG, The Hague, The Netherlands; email:
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
