The degree to which academic freedom is protected and respected in democracies tends to be relatively stable, and academic freedom tends to thrive under democracies (Berggren and Bjørnskov Reference Berggren and Bjørnskov2022). However, there are times when, even in democracies, academic freedom is under attack. Some governments are now targeting academic institutions such as the Central European University in Hungary, Boğaziçi University in Turkey, or the Center for Research and Teaching in Economics (Centro de Investigación y Docencia Económicas; CIDE) in Mexico. In April 2025, hundreds of university presidents in the United States signed a letter denouncing the “unprecedented government overreach and political interference” that higher education is experiencing under the Trump administration.Footnote 1 What explains attacks on academic freedom among electoral and liberal democracies? We argue that part of the explanation is the nature of political leadership. Populist leaders are naturally inclined to attack academic freedom because it conflicts directly with their values and worldview. As their time in office increases and their policies and politics are scrutinized by public opinion, pundits, and experts, populist leaders become less tolerant of critical voices. Academics, being part of these dissenting voices, become one of the obvious targets of populist leaders.
Mudde (Reference Mudde2004) defines populism as an ideology that views society as divided into two opposing groups—“the pure people” and “the corrupt elite”—and claims that politics should reflect the general “will of the people.” Based on this definition, Mudde argues that two values—elitism and pluralism—conflict with populism. Academic institutions embody these opposing values because their high degree of specialization and freedom of thought and expression necessarily precondition their dedication to fostering knowledge, research, and learning, which by definition cannot be univocal but are plural. Given their lack of tolerance of dissent to their governance, populist leaders are also intolerant of any critical thinking that questions or critically evaluates their governing plans, programs, and style. Because they have been voted in democratically, they tend to justify their attacks on academic freedom as necessary measures to represent the general will of the people who put them in power.
Although common wisdom associates populism with autocratic rule, we can observe that even in democratic contexts, which rely on freedom of information, academic freedom deteriorates under populist leaders. Using data from the Global Populism Database and controlling for factors such as constitutional protections of academic freedom, polarization, and legislative and judicial constraints on the executive (as included in the V-Dem dataset by Coppedge et al. Reference Coppedge, Gerring, Knutsen, Lindberg, Teorell, Altman, Bernhard, Agnes Cornell and Gastaldi2023), we find that as a populist leader’s rhetoric becomes more extreme, academic freedom significantly decreases. This decline persists even when accounting for a country’s level of liberal democracy, which is supposed to mitigate the harmful effects of populism on democracy (Mudde and Rovira Kaltwasser Reference Mudde, Kaltwasser, Mudde and Kaltwasser2012; Ruth-Lovell, Lührmann, and Grahn Reference Ruth-Lovell, Lührmann and Grahn2019). In-depth accounts of the attacks on academic communities in Mexico illustrate how academic institutions naturally become targets of populist rhetoric and how their academic freedoms suffer for it. They are portrayed as aligned with elite privilege, yet they also embody critical thinking and pluralism, which stand in direct opposition to the Manichean worldview advanced by populist leaders. As a result, they experience attacks on their academic freedoms evidenced by budget cuts, changes in leadership and personnel demotions, and the elimination of research initiatives and programs that were critical of the populist agenda.
This study makes three contributions. First, it contributes to the emerging comparative literature on academic freedom and its relationship to the tensions between liberal democracy and populism. Our study is among the first large-N comparative analysis examining the effects of liberal democratic institutional safeguards on the protection of academic freedom. However, unlike the studies by Berggren and Bjørnskov (Reference Berggren and Bjørnskov2022) and Lott (Reference Lott2024), which focus on the differential impact of autocracies versus that of democracies, we analyze the impact of populism on academic freedom. Second, by focusing on populism as the primary explanatory variable, we find that populist attacks on academic freedom have a negative effect even within the context of liberal democratic institutional settings. The negative impact of populist rhetoric on the protection of academic freedom in countries with constitutional safeguards and legislative and judicial constraints on executive power highlights that populist discourse not only reflects the autocratic tendencies of the leader but also increases populism’s inherent intolerance of criticism. Our findings demonstrate that populist leaders’ inability to tolerate the critical thinking of academic communities persists even in the presence of robust liberal democratic institutions. Finally, by providing an account of attacks against the CIDE in Mexico, one of the most prominent academic institutions in Latin America, we contribute to the expanding literature on the prevalent threats to academic freedom around the world.
Academic Freedom
Academic freedom can be defined as the “personal liberty to pursue investigation, research, teaching, and publication of any subject as a matter of professional interest without vocational jeopardy or threat of other sanction, save only upon adequate demonstration of an inexcusable breach of professional ethics in the exercise of that freedom” (Van Alstyne Reference Van Alstyne and Pincoffs1975, 71). It is essential for fostering intellectual curiosity, innovation, and the pursuit of truth within educational institutions. Academic freedom also serves as a cornerstone of democratic societies, enabling the free exchange of ideas, the criticism of arguments and policies based on their merits and the evidence available, and the advancement of scholarship for the betterment of society.
Some of the most powerful arguments for defending academic freedom have to do with its role in the pursuit of truth for the betterment of society and its ability to educate citizens so they can engage in a rational exchange of reasons while respecting other persons, both of which have a positive impact on democratic practices. The pursuit of truth argument derives from John Stuart Mill’s idea that the most effective method for expanding our repository of true beliefs is by fostering a free marketplace of ideas in which no opinion is excluded (Lynch Reference Lynch and Lackey2018). Practices of the academic profession, such as rational inquiry and the use of the scientific method, require critically evaluating the propositions presented and “following reason and method wherever they lead” (Andreescu Reference Andreescu2009, 502). Therefore, professors should be protected from attempts by the state or private actors to alter their research and teaching for political reasons. It also has been argued that governments grant universities autonomy and freedom in recognition of their social contributions. These contributions include preparing students for high-skilled employment, generating knowledge, and fostering innovations that drive economic growth. Despite potential criticism, this autonomy remains upheld when universities are perceived as continuing to serve vital roles in social progress (Cole Reference Cole2017).
Academic freedom, particularly when related to teaching in universities, also plays a social mission by contributing to the education of critical citizens engaged in public deliberation. Universities promote rational inquiry and the exchange of reason while respecting the opinions of others. Academic practices also train students and professors to develop critical thinking in their professional endeavors, which could be applied to other tasks in daily life. Finally, academic practices “serve as an engine for dissent. Dissent is vital in a democracy,… because it is through dissent that diversity of opinions and minority viewpoints can be heard” (Lynch Reference Lynch and Lackey2018, 33). If democracies aspire to be spaces of reason, then academic freedom must be protected and cherished.
However, some governments are targeting academic institutions such as the Central European University in Hungary, Boğaziçi University in Turkey, the University of Florida, or the Central American University in Nicaragua. The extent of protection of academic freedom has fluctuated throughout history, and it often faces its greatest challenges under authoritarian regimes. Such governments attempt to impose a unified view of society aligned with their ideologies and agendas, leaving universities particularly vulnerable to ideological control. For instance, under the latest military government in Brazil, higher education programs were subordinated to national development plans. Governmental bodies like the Federal Council of Education gained the authority to intervene in universities, influencing the appointment and promotion of staff, student admissions, course content, and budget management. These interventions resulted in inefficiency and a reduction in academic quality (De Figueiredo-Cowen Reference De Figueiredo-Cowen2002).
There are many ways to infringe on academic freedom, and even in democracies it can be under attack. For example, in the United States, legislation was passed during the McCarthy era requiring professors to sign loyalty oaths and to testify their ideological commitments, President Nixon had a list of enemies, and President Reagan did what he could to eliminate funding for the social and behavioral sciences. These measures discouraged open discourse and debate (Cole Reference Cole2017). Robert Hutchins, president of the University of Chicago, said this about intimidatory practices during McCarthyism: “The question is not how many professors have been fired for their beliefs, but how many think they might be. The entire teaching profession is intimidated” (as cited in Cole Reference Cole2017, 864).
Therefore, even in democracies, the level of protection afforded to academic freedom can vary over time. However, we still lack systematic empirical evidence that explains this variation. To fill this vacuum, Berggren and Bjørnskov (Reference Berggren and Bjørnskov2022) conducted an exploratory study of the variations in academic freedom across 64 countries. They found that democratization, even a minimalist electoral democracy approach, has a positive impact on academic freedom, whereas communist regimes exhibit less academic freedom. In terms of institutional design, judicial accountability supports academic freedom; turning to political variables, party concentration in the legislature affects that support, but this impact is substantially more negative in autocracies.
Academic Freedom and Populist Leaders
We argue that because populist leaders use a narrative that simplifies complex issues, rely on charismatic leadership, and advocate for policies that resonate with popular sentiment, universities and academics become a natural target in their attempt to promote a simplistic and Manichean understanding of the world and to control the flow of information to shape public opinion in their favor. Populist leaders may perceive academic institutions as bastions of liberal or progressive values that they oppose. By targeting academic freedom, they can wage a form of cultural warfare against these values and attempt to reshape societal norms in line with their own worldview. Universities are not only characterized by a pluralism of ideas but also have an elitist character: these two attributes—pluralism and elitism—are in direct conflict with the values and vision of populist leaders.
Populism can be described as a “thin ideology” that “considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’” and emphasizes that the popular will should prevail (Mudde Reference Mudde2004, 543). Particularly important for our argument, this definition of populism “has two opposites: elitism and pluralism” (543). Universities could be considered elitist institutions because they tend to have a reputation for academic excellence and prestige. By requiring a high degree of specialization and academic credentials, the academic sphere sets itself apart from the skills of common people. Additionally, many universities have selective admission processes based on academic performance, which makes them accessible only to a selected minority. These factors could contribute to their perceived elitism.
Additionally, universities provide a fertile ground for the cultivation of pluralism by encouraging professors and students to explore and debate diverse viewpoints without fear of censorship, guided by principles of rational inquiry and respect for others’ opinions. Furthermore, academic practices help professors and students develop critical thinking skills, making them a natural engine of dissent and informed critique (Lynch Reference Lynch and Lackey2018; Rabban Reference Rabban2024). Pluralism, dissent, and critique are values that come into conflict with populist leaders who aim to control public discourse and reduce it to simplistic solutions derived from a dramatized battle between good and bad, without nuance and a rationale. That is why research universities and the academic freedom that protects their work become a clear target of populist leaders. In the 1960s, a time of social unrest in the United States, the University of Chicago commissioned a committee to elucidate the role of universities during difficult times. In 1967, the Kalven Committee concluded, “A university faithful to its mission will provide enduring challenges to social values, policies, practices, and institutions. By design and by effect, it is the institution that creates discontent with the existing social arrangements and proposes new ones. In brief, a good university, like Socrates, will be upsetting.”Footnote 2
If academic freedom is not protected, open debate, dissent, and critique are at risk, leading to negative consequences for society: “Without a university that is permitted … to reign free and despite harsh criticism of government policies, by faculty and students, to be supported by the government, then society has lost its democratic soul” (Cole Reference Cole2017, 864). Given that the core concepts of populism—the people, the elite, and the general will—are empty categories, it is the populist leader who gives them meaning, determining who qualifies as “the people” and who constitutes “the elite” (Mudde and Kaltwasser Reference Mudde, Kaltwasser, Mudde and Kaltwasser2012). As a result, the inherent tension between populism, a monistic ideology that claims to represent the will of the people, and academia, which promotes pluralism, may lead a populist leader to attempt to delegitimize academics. By portraying them as an elite disconnected from the people’s interests, the leader can position them as adversaries of the people.
Populism can also be understood as a political strategy in which “a personalistic leader seeks or exercises government power based on direct, unmediated, noninstitutionalized support from large numbers of mostly unorganized followers” (Weyland Reference Weyland2001, 14). For this strategy of direct mobilization to work, the leader relies on charisma as a force that allows him to subordinate the people to “the leader’s personal will.” This implies a mobilization strategy that primarily depends on emotion, rather than reason. Thus, the main basis of power of populist leaders relies on “the plebiscitarian backing of citizens impressed by their exceptional personal qualities and convinced of their magical powers” (Weyland Reference Weyland2025, 197). However, when these leaders’ performance makes it difficult to sustain the narrative of extraordinary accomplishments, they may seek to preserve authority by manipulating the political discourse and controlling the messages that citizens receive (Guriev and Treisman Reference Guriev and Treisman2022). Thus, in the process of subordinating the masses to the will of leaders—based on the belief that they alone represent “the people”—pluralism, dissent, and critique come to be seen as obstacles that must be eliminated because they hinder implementation of the populist leaders’ political agenda. Knowledge institutions become the control obsession of populist leaders because of their inability to tolerate pluralism, dissent, and critical thinking—the three inherent characteristics of learning and innovation (Jackson Reference Jackson2019). Attacking academic freedom, therefore, can be viewed as an ideal tool to control both the discourse and the political agenda.
Populist leaders should therefore be naturally inclined to attack academic freedom because it conflicts directly with their values and worldview and because it is a strategy that facilitates their manipulation of the messages that people receive.
Quantitative Empirical Analysis
We examine this hypothesis using a cross-sectional time-series analysis of electoral and liberal democracies around the world between 2000 and 2021. We focus on this period because it coincides with the global rise of populist governments and allows us to examine recent trends in attacks on academic freedom. Although this wave began in the 1990s, it notably intensified during the 2000s, reaching an all-time high in 2018 (Funke, Schularick, and Trebesch Reference Funke, Schularick and Trebesch2023). Our dataset covers 59 countries, aiming to include every electoral and liberal democracy, but it is limited by the availability of information for measuring the populist discourse of political leaders (Global Populism Database v2; Hawkins et al. Reference Hawkins, Aguilar, Silva, Jenne, Bojana Kocijan and Kaltwasser2022). We selected our cases using V-Dem’s “Regimes of the World - The RoW Measure” (v2x_regime), which provides a robust classification of regime type, and restricted our analysis to country-years with a value of 2 (electoral democracy) or 3 (liberal democracy).
Our main dependent variable is the V-Dem’s index of academic freedom (v2xca_academ), which is measured on an interval scale ranging from 0 to 1. This index is constructed based on various indicators, including freedom to research and teach, freedom of academic exchange and dissemination, institutional autonomy, campus integrity, and freedom of academic and cultural expression. Additionally, as robustness checks, we use other indicators from V-Dem as alternative measures of the dependent variable, such as the freedom to research and teach (v2cafres) and freedom of academic exchange and dissemination (v2cafexch).
The data generally reveal a trend of stability and respect for academic freedoms, as expected in democratic countries. However, there are exceptions indicating a consistent and ongoing decrease in academic freedom levels, which was notably observed in Hungary, India, Nicaragua, and Turkey. Other countries experienced a more abrupt decline in academic freedom, as exemplified by Brazil, followed by Bolivia and Mexico, and some countries such as Colombia were characterized by a persistent state of relative weakness in academic freedom. Figure 1 illustrates these dynamics of the academic freedom index for the country-years within our sample for which data on the independent variable are available.
Visualization of the Academic Freedom Index for Democratic Country-Years in which the Populism Index Is Available
Note: Darker colors represent higher levels of academic freedom. Source: Authors’ elaboration with V-Dem data and the Global Populism Database v2.
Our main independent variable attempts to capture the presence of populist leaders by measuring the populist discourse of political leaders based on data from the Global Populism Database (Hawkins et al. Reference Hawkins, Aguilar, Silva, Jenne, Bojana Kocijan and Kaltwasser2022). Hawkins and his team of researchers codify discourses of prime ministers and presidents for each term in office by analyzing four categories of their speeches: a campaign speech, a ribbon-cutting speech, an international speech, and a famous speech. Speeches are graded from 0 to 2, where 2 represents a “speech [that] expresses all or nearly all of the elements of ideal populist discourse” and 0 represents speeches that use “few if any populist elements. Note that even if a speech expresses a Manichean worldview, it is not considered populist if it lacks some notion of a popular will.” This measure applies to populists on the left and the right. Across the country-years included in the analysis, the average value of populist discourse is 0.26 (s.d. = 0.33); the minimum value is 0, and the maximum value is 1.73. Note that this variable was measured at the level of presidential or parliamentary administrations, covering entire terms rather than individual years. Because yearly measures of populism were unavailable, we assigned the same value for each administration to all the years within its term to enable a yearly analysis (Hernández-Huerta and Inclán Reference Hernández-Huerta and Inclán2025). Figure A1 in the appendix shows a visualization of this variable for the countries in our sample across time.
We included control variables that could likely be associated with academic freedom. Following Berggren and Bjørnskov (Reference Berggren and Bjørnskov2022), we first included some political and institutional characteristics such as judicial and legislative constraints on the executive, which might limit the government’s ability to attack academic freedoms. Instead of adding dummies to differentiate between electoral autocracies and democracies—and given that our study is restricted to electoral and liberal democracies—we included V-Dem’s index of liberal democracy to account for their finding that democratization is positive for academic freedom. In addition, following Berggren and Bjørnskov (Reference Berggren and Bjørnskov2022), we included controls for key economic indicators such as GDP per capita, inflation, and the logarithm of population size. We also used a measure of the constitutional provisions for the protection of academic freedom on the assumption that these provisions could safeguard this principle, shielding academic institutions from political interference and ensuring stability across administrations. Finally, we added a measure of the degree of polarization in a society because it may lead to increased pressure on academic institutions to conform to political ideologies, potentially limiting the scope for independent research and the expression of diverse viewpoints.
For our estimation strategy we decided to use a fixed-effects linear regression analysis because our data have a cross-sectional time-series structure with a continuous dependent variable. This method allows for the control of country-specific characteristics that remain constant over time, thereby reducing bias and improving the precision of estimates. Additionally, fixed-effects models provide insights into within-country changes over time, offering robustness against omitted variable bias.
Table 1 presents our findings. Model 1 presents a bivariate model in which we estimate the direct relationship between our measure of populism and the index of academic freedom, with country-fixed effects. Model 2 presents estimates of the previous relationship but adds the whole set of controls, except for the liberal democracy index. Model 3 presents the full model with the index of liberal democracy included. Finally, model 4 presents the full model minus the economic controls, GDP per capita, and inflation, because adding these controls reduced the sample size by about 400 observations, almost half our sample.
Fixed-effects Linear Regression Models for Academic Freedom
Note: Standard errors in parentheses.
*** p < 0.01, ** p < 0.05, * p < 0.10.
Because the concept and measure of liberal democracy might be highly correlated with other controls, we did a variance inflation factor (VIF) test to detect a potential problem of multicollinearity. The VIF value for this variable was 6.24, which indicates moderate multicollinearity, suggesting that there may be some degree of correlation between the liberal democracy index and other independent variables in the regression model. Although there is no strict cutoff for determining when multicollinearity becomes problematic, a general guideline is that VIF values above 5 or 10 may indicate significant multicollinearity. All other variables have a VIF value lower than 5. This is why we present separate models with and without the index of liberal democracy.
All four models in table 1 indicate that leaders with higher levels of populist discourse exhibit a statistically significant negative relationship (p < 0.01) with academic freedom. These findings suggest that as leaders adopt more populist rhetoric, academic freedom tends to decline. To examine the substantive effects, we plot the predicted probabilities for the level of academic freedom over the range of a leader’s populist discourse (figure 2). This prediction is based on model 2, with control variables held at their means and the only categorical variable, constitutional protections, at its modal value. As shown here, populist presidential discourse corresponds with decreases in academic freedom. When populist discourse is equal to 0.2, next to the minimum observed value, the estimated academic freedom is at 0.8885 (in a 0–1 range). Then, as populist discourse increases, the level of academic freedom declines. When populist discourse is at its maximum observed value, 1.73, the level of academic freedom declines to 0.8643. To contextualize this change, a shift from the lowest to the highest level of populist discourse corresponds to a decline in academic freedom similar to moving from the level of the United Kingdom (2017–18) to that of Serbia (2004–8) or Ecuador (2000–5).
Predicted Probabilities of the Level of Academic Freedom across Changes in Populist Discourse
Note: A histogram shows the distribution of populism scores across their full range.
Although the decline in the magnitude of the coefficient may seem moderate, it is not negligible, considering that the outcome is measured on a 0–1 scale and given the remarkable stability and good standing of academic freedom in democracies. In figure 1 we embedded a histogram illustrating the frequency of populism scores across their possible range. As expected, most leaders in electoral and liberal democracies exhibit low levels of populism. However, we believe that the fact that the statistical analysis still identifies a significant relationship between populism and academic freedom, despite the relative rarity of highly populist leaders in democracies, provides strong evidence of a meaningful association between the two variables.
We also find that, in line with previous findings (Berggren and Bjørnskov Reference Berggren and Bjørnskov2022), democracy significantly influences academic freedom. In model 3, the estimate of the liberal democracy index shows a positive and statistically significant relationship (p < 0.01) with the level of academic freedom. This indicates that in societies with higher levels of liberal democracy, academic freedom tends to be more robust. Essentially, the foundational aspects of liberal democracy, such as the protection of individual and minority rights from state and majority oppression, in conjunction with the institutions to protect those rights, such as a strong rule of law and an independent judiciary, collectively contribute to fostering an environment where academic freedom thrives. This underscores the critical role of democratic principles in preserving and promoting academic freedom. In figure 3, we plot the predicted probabilities of academic freedom based on model 3’s estimates of changing values for liberal democracy. Going from the lowest observed value of liberal democracy in our dataset, 0.268, to the highest level of 0.896 results in an increase of 0.18 points in the index of academic freedom, which represents almost one-fifth of the potential total range of the index. However, as we mentioned earlier, the interpretation of results in model 3 (table 1) should be approached with caution because the model may experience moderate multicollinearity when the liberal democracy variable is included.
Predicted Probabilities of the Level of Academic Freedom across Changes in the Liberal Democracy Index.
A particularly interesting feature of these findings is that, when analyzing the impact of populism on some phenomenon, it can be challenging to distinguish between the authoritarian tendencies of political leaders and their populist inclinations, because these concepts often overlap. In model 3 we introduce liberal democracy as an alternative control variable. By interpreting liberal democracy as the antithesis of authoritarianism, we can suggest that the adverse effect of populism on academic freedom is independent from that of an authoritarian context characterized by restricted political pluralism, limited freedom of speech and expression, curtailed civil liberties, lack of an independent judiciary, and minimal checks and balances on government power. Returning to the impact of populist leadership, our main independent variable, its statistical significance and coefficient size, even when its coefficient size seems moderate, appear to be unaffected by the inclusion or exclusion of the liberal democracy control. We suggest that these results should be interpreted as evidence of an additional effect that populist discourse has on academic freedom—one that operates independently of the important and broader impact of a regime’s democratic or authoritarian nature (Berggren and Bjørnskov Reference Berggren and Bjørnskov2022; Lott Reference Lott2024), as captured in the liberal democracy index.
Lastly, we turn our attention to some control variables. Our indicator of societal polarization ranges from 0, indicating significant polarization, to 4, denoting no polarization. Therefore, we would expect to see a positive relationship between this variable and academic freedom. As expected, models 2 and 3 confirm this relationship, showing a positive and statistically significant association between the societal polarization indicator and academic freedom. This suggests that academic freedom is inhibited in more polarized contexts. Meanwhile, the variables measuring legislative and judicial constraints on the executive do not provide conclusive or consistent evidence of a relationship with academic freedom.
As part of our robustness analysis, we replicated the same models using alternative measures of our dependent variable. Instead of using V-Dem’s index of academic freedom, we employed alternative measures of the dependent variable such as the freedom to research and teach and the freedom of academic exchange and dissemination. Both variables are rated on a scale from 0 to 4, where 0 indicates complete restriction of the respective freedom and 4 signifies full freedom. The first alternative variable attempts to capture the extent to which scholars can freely develop and pursue their research and teaching agendas without interference, and the second variable aims to capture the freedom that scholars enjoy toexchange and communicate research ideas and findings. Using these alternative measures of the dependent variable, the coefficient for populist leadership is six to seven times larger than the results obtained with the academic freedom index. Additionally, the populism leadership variable remains statistically significant at the p < 0.01 level across all eight models (see tables a2 and a3 in the appendix). Table A4 presents an additional robustness check. It shows that even when using an alternative measure of regime type for case selection—specifically, the Boix et al. regime measure (e-boix-regime) instead of v2x_regime—our results remain consistent: the populist variable continues to have a negative sign and remains statistically significant at the p < 0.01 level.
In addition, we restricted our analysis to events occurring after 2000, because our goal is to focus on recent cases worldwide. To ensure that our results are not sensitive to this cutoff, we replicated the regressions presented in table 1, this time including the available data for cases before 2000. The results remain largely unchanged and are reported in table A5 as an additional robustness check.
Finally, to account for the possibility that populist leadership may not have an immediate impact on academic freedom and as an additional robustness check, we reran the analyses presented in table 1 using different lags of the populism variable. Specifically, we included a one-year lag (populism at t − 1 affecting academic freedom at t) and created a new variable representing the average populism value over the two prior years, excluding the current year (average of t − 1 and t − 2 affecting t).
The results of these analyses are presented in tables A6 and A7 in the appendix. In these models, the size of the populism coefficient decreases by less than half and remains statistically significant at the p < 0.01 level in half the models. In the cases where it lost statistical significance, this may be due to the reduction in sample size. When we removed the economic controls—which had nearly halved the sample size—the populism variable regained significance. Overall, tables A6 and A7 show a slight reduction in both effect size and significance. However, it is important to note that the populism variable is measured by the administration of the leader, rather than annually. This measurement approach limits the precision of estimating yearly changes and may dilute the observed effects of the lagged variables.
The Case of CIDE in Mexico
The case of the Center for Research and Teaching in Economics (CIDE), in Mexico illustrates the differentiated patterns of academic attacks that populist leaders may implement. Where constitutional protections for academic freedom exist (Ginsburg Reference Ginsburg2022; Rabban Reference Rabban2024), populist leaders turn to discursive, budgetary, and political attacks aimed at (1) undermining the institution’s prestige and capabilities and (2) replacing leadership and driving away personnel. These actions can be seen as part of a broader populist strategy to promote a Manichean worldview in which academics are portrayed as part of a corrupt elite. By seeking to control the flow of information and belittling external and independent sources of information, populist leaders can construct the idea of a homogeneous general will that stands in sharp contrast to the logic of deliberation and pluralism that characterizes knowledge institutions (Jackson Reference Jackson2019).
In 2018, Andrés Manuel López Obrador (AMLO) came to power with a strong electoral mandate, capitalizing on widespread social discontent and a crisis of representation among traditional political parties.Footnote 3 He had been characterized as a leader whose discourse aligns with that of a prototypical populist, featuring a narrative of people-centrism and anti-elitism and framing politics as a Manichean confrontation between good and evil (Dussauge-Laguna Reference Dussauge-Laguna2022; Monsiváis-Carillo 2023). Under the auspices of corruption eradication, he began his term by dismantling or colonizing public agencies and autonomous watchdog organizations that posed accountability limitations to his efforts to concentrate power under one-man rule (Ibarra del Cueto Reference Ibarra del Cueto2023; Monsiváis-Carillo Reference Monsiváis-Carrillo2023). These practices, such as weakening horizontal checks and balances, are tools commonly used by populists to contribute to democratic backsliding. Other tactics for undermining political accountability include attacking and discrediting civil society and independent media (Prushankin and Kaltwasser Reference Prushankin, Kaltwasser, Croissant and Tomini2024). Populist leaders use their rhetorically self-proclaimed monopoly on the “will of the people” to legitimize policy decisions that diminish the importance of experts and scientific evidence (Dussauge-Laguna Reference Dussauge-Laguna2022). In this context, academic institutions became natural targets of his populist rhetoric, because they symbolized the elites’ privilege and hotbeds of corruption, against which he needed to fight as a representative of the will of the virtuous people who put him in power. Academic communities also became his preferred targets, given their characteristic critical thinking that prevented them from offering conceding opinions. They simply could not ignore their responsibility to evaluate social and economic policies and so became critical voices of AMLO’s political and policy decisions. Doing their expected job put them at cross-purposes with the president’s populist agenda.
Discursive Attacks
The attacks on CIDE by AMLO’s administration should be understood within a broader context. As is typical of a populist leader seeking to control the public narrative, AMLO showed disdain for science and any other sources of information that might discredit his views. There are several examples of AMLO deriding science. For instance, in the midst of discussions on the 2013 Energy Reform bill, which allowed private and foreign investment across the energy values chain, AMLO claimed, “It is not a matter of us lacking technology, you shouldn’t believe that it is a lot of science to drill an oil well, because it is as if drilling a water well, the only difference is that in the water well the depth is 25, 30 meters and in the oil sector they are 3,000 to 5,000 meters deep.”Footnote 4 In his effort to weaken the federal public administration and ridicule bureaucrats by firing them en masse, AMLO stated, “Don’t think that governing has much science; … politics has more to do with common sense,”Footnote 5 revealing his belief that knowledge and preparation are not essential for government decision making or public policy design, but that intuition is more important. During the COVID-19 crisis, AMLO issued public health recommendations that were not aligned with evidence. Rather than promoting social distancing, he urged Mexicans to attend parties, dine in restaurants, and go shopping. He also regularly appeared at public events without wearing a facemask and questioning the value of expertise (Knaul et al. Reference Knaul, Arreola-Ornelas, Porteny, Touchton, Sánchez-Talanquer, Méndez and Chertorivski2021). He even stated that his best protection against the coronavirus was provided by his honesty and the pair of religious images he carried in his wallet, saying, “These are my bodyguards against the coronavirus … they will protect me.”Footnote 6 These discursive strategies may have been an attempt to empathize with uneducated masses to maximize political gain, rather than to achieve policy outcomes, but they also showed a clear disdain for expertise.
It was in this context of disputing the value of knowledge and scientific evidence that AMLO began publicly attacking the reputation of academic institutions that dared to criticize his handling of the pandemic, his security strategy, and his infrastructure projects; these attacks laid the groundwork for subsequent budgetary and administrative assaults. In December 2018, shortly after taking office, López Obrador sent an education bill to Congress that made no mention of the autonomous status that public higher education institutions had historically enjoyed. After the opposition raised concerns about this omission, AMLO clarified during his daily press conference that there was no intention to eliminate the autonomy of higher education institutions and that the bill would be corrected. However, since then, he did not miss any opportunity to criticize the National Autonomous University of Mexico (Universidad Nacional Autónoma de México, UNAM), accusing it of moving to the right and becoming gentrified (Velázquez and Baranda Reference Velázquez and Baranda2021). Other public academic institutions also faced discretionary leadership changes, budget cuts, and discursive attacks.
Yet CIDE became the primary target of López Obrador’s hostility because he wanted to set an example for the remaining 26 public research centers under the authority of the National Council of Science and Technology. These institutions were expected to act in alignment with the federal government and to refrain from critically assessing its public policies.Footnote 7 Although other academic institutions also came under attack, they did not experience the same level of hostility—either because their budgets were relatively protected by constitutionally guaranteed autonomy, as in the case of UNAM, or by autonomy granted by executive order, as with El Colegio de México (COLMEX), or they were financially independent from federal allocations, as were private institutions like ITAM.Footnote 8
CIDE was founded in 1974 as a public research center specializing in the teaching and research of economics. Since then, its faculty have provided teaching, research, and advice on political, economic, and legal matters regarding public policy because doing so is inherent to its mission. CIDE is one of the most prestigious research centers for social sciences in Mexico, gaining a world-class reputation since the 1990s. Its professors publish in top international academic journals and are sought out as experts by the media, NGOs, the Mexican government, and international institutions. Its research has informed public policy in Mexico, making it a significant player in the country’s academic and political landscape.
From the early days of his administration, López Obrador sought to portray CIDE as an elitist institution disconnected from the interests of the people. Its production of independent research that questions and even criticizes government’s policies and outcomes achieved by AMLO’s and previous governments may explain why a populist leader might see CIDE as a political rival. AMLO earned a reputation for dismissing statistics prepared by independent agencies that contradict his narrative by claiming, “I have different data.” In this struggle over facts and the representation of truth, an institution like CIDE could become an obstacle to a leader who seeks direct and unmediated communication with society, aiming to monopolize the agenda and control the facts communicated to his followers.
In his discursive attack on CIDE, Mexico’s president criticized it for becoming a center for training private sector cadres with a neoliberal ideology, implying that it had abandoned its mission of public service. In August 2021 he stated, “I understand that ITAM exists, an institution financed by the private sector to train its cadres, but not that the State is financing an institution with those same purposes such as the CIDE…. They are telling young people, training them to listen to the [International] Monetary Fund.”Footnote 9 This rhetoric aimed to discredit CIDE by portraying it as an elitist institution aligned with neoliberal ideologies, creating a narrative that resonated with AMLO’s base.Footnote 10
After the early resignation of the center’s director and the federal government’s appointment of an outsider, students and professors protested what they considered to be an illegal imposition that transgressed CIDE’s norms and collegial decision making. In response, on November 29, 2021, AMLO asked reporters if they were blaming him for this leadership change. He said, “This institute is interesting; it was created to train economists or technicians to aid public administration … but then it embraced neoliberalism and started competing with ITAM…. The recent directors of CIDE embraced this orientation. They all supported neoliberal policies and remained silent about the looting that occurred over 36 years.” He, then, concluded that this institution had shifted to the right, and its leaders “weren’t up to the task in the face of the greatest looting in Mexico’s history. A group of intellectuals, very conservative and accommodating, took no risks to climb the social ladder, remaining complicitly silent.”Footnote 11 These statements aimed to blame CIDE officials and professors for the country’s past problems and to distance them from the public by portraying them as part of a conservative elite complicit in looting the country, whose research did not contribute to Mexico’s well-being.
One week later, Mauricio Merino, a former CIDE professor, replied to AMLO’s accusations:
The smear campaign I am talking about is personified and sustained by Andrés Manuel López Obrador, the President…. He once again argued that we had remained silent like mummies in the face of the looting and corruption of previous governments…. What President López Obrador has said is not true, and it must be stated. As part of the defense of democracy, we must set the truth straight and not tolerate lies and slander from public authorities.
To make his point, Merino showcased a large number of books and studies published by CIDE faculty. He called attention to a study on the structure of accountability in Mexico, which was the initial driving force for discussions on the need to create a national anticorruption system. He pointed to books on public ethics and open government, highlighting the lack of transparency of previous governments. Together with the Red por la rendición de cuentas, a Mexican NGO, CIDE researchers drafted the current General Transparency Law and designed the National Transparency System. Merino presented the Procampo denunciation made by CIDE, along with FUNDAR, a think tank in Mexico, and other organizations, documenting how state and federal governments were stealing money under the pretense of helping the agricultural sector. He also showcased books on the problems of Mexican democracy and the need to modify public policies to prevent corruption, emphasizing CIDE’s recognition of the needs of the poorest people. Merino also cited CIDE’s studies demonstrating bias in the allocation of public resources and the need to defend and improve the country’s social programs. After presenting these and other examples of how CIDE’s research contributed to the nation’s well-being, Professor Merino concluded, “We did report and confront the corruption of all governments since the 1980s and even before, without exception. We have fought for the establishment of institutions to combat opacity and the lack of accountability. And let it be known, we will continue to do so with this government, just as we did with the previous ones, despite the presidential fury.”Footnote 12
But the president did not stop his attacks on CIDE. A few months later, AMLO reiterated, “Unfortunately, what they did with CIDE was to make it a version very similar to ITAM, and even a second-rate one, because ITAM had teachers like Pedro Aspe, José Antonio Meade, and Luis Videgaray—big names in conservative thought. CIDE became a copy, a second-rate version. It didn’t start that way, as it was intended to train public servants for the nation’s interests, but it ended up despising public service.”Footnote 13 These attacks on the center’s reputation were not surprising, given the president’s demonstrated tendency to alter reality in his public statements. During his first year in office, 538 of his verifiable statements were analyzed, and 45% were found to be misleading or false.Footnote 14 In the process of politically appropriating the truth, the discursive attacks served as a strategy to discredit the institution in the eyes of the public. For a populist leader, this was a calculated strategy because independent research from a credible source could disrupt the presidential narrative, preventing it from monopolizing public discourse and strengthening his personalist connection with the masses (Jackson Reference Jackson2019). To back up his statements with actions, the populist needs to impose instrumental actors on the institution who will enact his vision—in this case, the appointment of leaders of public research centers with the clear mandate to limit academic freedom.
During these attacks, CIDE received public support from domestic and international universities and research centers. Both individual and institutional letters were published on social media, but the discursive attacks intensified. María Elena Álvarez-Buylla, director of Consejo Nacional de Ciencia y Tecnología (CONACYT)—Mexico’s equivalent agency to the US National Science Foundation, compared CIDE and several of its professors and administrators to a criminal cartel during one of the president’s daily press conferences, speaking with the president standing behind her. In 2023, a public information request forced the authorities to acknowledge that there was no evidence to support these claims.Footnote 15 However, no apology, retraction, or sanction followed. As we describe in the next section, there were also budgetary attacks on CIDE, resulting in severe consequences for students, professors, and administrators.
Budgetary Attacks
The government under AMLO enacted two budgetary measures that severely impaired the functioning of CIDE: its budget was slashed, and a trust fund that enabled the institution to generate its own funds was eliminated. In October 2020, in an eventful legislative session, the left-wing MORENA Party, with the support of its allies, the Labor and Green Ecologist Parties, passed a measure to allow the executive to eliminate and concentrate 109 trust funds that had been created to finance not only science and technology but also provide support for victims of criminal and domestic violence, journalists, and human rights defenders at risk; emergency funds in the wake of natural disasters; daycare services for working women; and health treatment and services for critically ill patients, among other services (Wade Reference Wade2019). The science and technology trust funds served as a financial mechanism to ensure consistent funding for specific programs, which annual government budgets, subject to political fluctuations, could not guarantee. They supported a range of programs from student scholarships and equipment maintenance to major research projects. Funding came mostly from international grants, donations, and the institutes’ own teaching activities. After these trust funds were eliminated, public research centers saw their research projects, multiannual budgets, and fundraising plans severely affected, as new funding mechanisms had to be put in place (Pérez Ortega and Gutiérrez Jaber Reference Pérez Ortega and Jaber2020). Without institutional programs to replace trust funds, administrators of public research institutions’ hands were tied, and researchers’ salaries were cut. Relations between sitting heads of public research institutions and CONACYT deteriorated as clear guidelines continued to be delayed, and personnel began to leave in search of better working conditions.
CIDE’s trust fund had provided a significant supplement to its federal funding, enabling a large number of researchers to engage in projects financed by various national and transnational public, private, and nonprofit agencies. A key objective of these projects was to enhance the design and implementation of public policy addressing cutting-edge social research topics. In the last 10 years of operation, CIDE’s trust fund generated an average annual income of 141.3 million pesos (approximately USD $7 million). Its lowest level of revenues was 91.5 million pesos in 2020, the first year of the COVID-19 pandemic, and the highest was 248 million pesos in 2017. It generated funds for an average of 90 new academic projects and continuing education programs per year.Footnote 16 Additionally, the funds were used for scholarships, to incentivize faculty research and teaching productivity, and to support faculty research travel, fieldwork, conferences, and network activities. With the elimination of the trust fund, the revenue it generated dropped to zero, and it did not support any projects in 2022, leading to a major operational challenge for the institution.Footnote 17
Budgetary changes for the ostensible purpose of increasing public access to higher education included the elimination of tuition fees. These moderate fees had allowed CIDE to offer merit-based scholarships to outstanding students and to those with financial needs.Footnote 18 Their elimination thus aligned CIDE’s admissions policy with López Obrador’s equalizing populist discourse. Without the support of the trust fund and without tuition income, CIDE could no longer afford to fund living expenses for students who, even without paying tuition, could not attend college because of financial challenges. Low-income talent was thus discouraged from pursuing a high-quality education at CIDE. As the number of scholarships decreased in recent years, the proportion of students from public high schools also declined, currently representing less than 30% of those admitted. CIDE had taken pride in being an instrument of social mobility. That social goal seems harder to achieve now.
Political Attacks
Political attacks against CIDE occurred in the hostile environment orchestrated by AMLO since the beginning of his administration through austerity measures that not only included budgetary cuts but also the elimination of scholarships, research grants, and public funding for research labs, programs, and centers; a controversial science reform bill; embezzlement charges against research that was contrary to the ideological point of view of CONACYT’s sitting administration and charges of organized crime and money laundering against 31 academics and administrators (Gutiérrez Jaber Reference Gutiérrez Jaber2021; Manjarrez Reference Manjarrez2023; Pérez Ortega and Gutiérrez Jaber Reference Pérez Ortega and Jaber2020; Reardon Reference Reardon2021; Wade Reference Wade2019).
In August 2021, CIDE’s director, resigned two years before the end of his second term (Miranda Reference Miranda2021). Other public research centers had already experienced similar resignations of their administrative personnel and the appointment of interim directors.Footnote 19 In the case of CIDE, CONACYT appointed an outsider as interim director with the intention of realigning the public research center’s research agenda with the president’s “industrial policies” and to educate students under a new paradigm, rather than its so-called neoliberal mission (Grupo Reforma 2021; Velázquez Reference Velázquez2021). In October 2021, the director of the campus in Aguascalientes was demoted because of a “lack of trust” after he publicly expressed support for 17 postdocs whose funding was to be ended by CONACYT (Camhaji Reference Camhaji2021). In November, CIDE’s academic provost was demoted because of her “rebellious behavior” after she defended the internal tenure reviews (Wong Reference Wong2021). In December, after CONACYT ratified the new director’s appointment, students took over the control of the center’s facilities by setting up encampments in both the Mexico City and Aguascalientes campuses.Footnote 20 In response, the coordinator of administration and finance and the outreach secretary offered their resignations (Redacción AN/MDS 2021). Students maintained their encampments for a month but removed them from both campuses in January 2022 to allow academic activities to resume after the winter break.
CIDE’s new leaders have shut down five interdisciplinary research units: the CLEAR Center, a World Bank-funded center for policy evaluation; the Drug Policy Program (PPD); the Interdisciplinary Program on Education Policy and Practices (PIPE); the National Public Policy Lab (LNPP); and the Program for the Study of Violence. Two master’s programs—one in research journalism and the other in environmental economics—have been eliminated. To justify these changes, CONACYT’s director, in one of the daily presidential press conferences, described the “networks of created interests at CIDE” as an example of a “neoliberal public research center” funneling resources away from the public good (Reardon Reference Reardon2021). In addition, the earlier mentioned contracts of the 17 postdocs assigned to CIDE through CONACYT’s Researchers for Mexico Program were terminated, and students and professors have experienced cuts in fellowships, scholarships, and research grants. More than 40% of CIDE personnel have resigned.
Conclusion
Our analysis suggests that, even after controlling for the protection of academic freedom granted by a liberal democracy, there is a strong negative association between the use of populist discourse by leaders and academic freedom, indicating that as leaders employ more populist rhetoric, academic freedom tends to decline. By analyzing the effects of populist discourse under democratic rather than autocratic settings, we show that the adverse effect of populism on academic freedom is independent from an authoritarian context characterized by restricted political pluralism, limited freedom of speech and expression, curtailed civil liberties, lack of an independent judiciary, and minimal checks and balances on government power. Hence, these findings confirm the hypothesis that guided this study; namely, that populist leaders will be inclined to attack academic freedom, which conflicts with their values and worldview and willingness to manipulate the messages that people receive.
This study makes three key contributions. First, we make a theoretical contribution by showing that there is an inherent conflict between populism as an ideology and characteristics commonly attributed to academic institutions—pluralism and an elitist nature. This tension between populism and academia makes academic institutions a natural and convenient target of populist rhetoric. In addition, populism, when conceived as a strategy that allows an unmediated connection between the leader and the masses, may lead populist leaders to attempt to control public discussion and manipulate the messages received by citizens, making academic institutions—which incentivize critical thinking and dissent—an obstacle to their control of the narrative they seek to impose. Second, we show there is a negative empirical relationship between the populist discourse of political leaders and academic freedom. Third, by documenting the attacks on CIDE in Mexico, we aim to contribute to the growing body of evidence on global threats to academic freedom.
In the aggregate, these contributions also suggest new directions for establishing a theoretical connection between the relationship between populism and democracy (Prushankin and Kaltwasser Reference Prushankin, Kaltwasser, Croissant and Tomini2024). Even if modern populist leaders typically do not use direct repression or coercion against their political opponents (Weyland Reference Weyland2025), they can inflict significant damage on liberal democracy by weakening some of its fundamental pillars. This is exemplified by their attacks on its core values, such as pluralism and dissent, via the attack on academic freedom. Attempts to impose a monist view of society, in which the leader defines who the enemy is and what constitutes the general will, may make societies collectively anesthetized, losing sight of the importance of deliberation, plurality, and dissent for a healthy and robust democracy. Viewed from this perspective, attacks on academic freedom could be viewed as one mechanism through which populist leaders embark on processes of democratic erosion.
The brief narrative of the attacks on CIDE in Mexico illustrates the causal rationale behind our arguments; in particular, the negative effects that populist discourse, materialized in budgetary and political decisions, can have on academic institutions. By purposely attacking CIDE, AMLO made clear his (mis)conception about what knowledge institutions were supposed to do, as well as his profound contempt for the work of academics. Through administrative and budgetary decisions, he severely weakened policy-evaluating initiatives, research projects, and academic programs. Through discursive attacks he also paved the way to control the flow of information, public opinion, and his personal connection with the public.
The case of CIDE also illustrates some vulnerabilities that this institution faced that could have been ameliorated through institutional measures, thereby offering valuable lessons for strengthening academic freedom. First, although our statistical models suggest that constitutional protections for academic freedom are not statistically significant in explaining variation in its levels, the case study of CIDE underscores a critical vulnerability: academic entities directly dependent on the national government, such as CIDE and other CONACYT-affiliated research centers, were particularly exposed to government interference. In contrast, public higher education institutions with budgetary and decision-making autonomy were less severely affected. This suggests that granting autonomy to all public higher education institutions can act as an essential safeguard for academic freedom by reducing their susceptibility to political interference. Second, the case study also reveals that CIDE was especially vulnerable to budgetary constraints imposed by the government, which were used to undermine its independence. Establishing legal provisions to protect the budgets of higher education institutions could serve as a mechanism to shield them from financial manipulation, particularly under populist leadership. These measures can create a more resilient framework to defend academic freedom against political interference.
At the onset of the new administration of President Sheinbaum, there was a glimmer of hope for academic freedom in Mexico. Despite MORENA winning the presidential and legislative elections by an even larger margin than the one that brought AMLO to power in 2018, the incoming president, Claudia Sheinbaum—a former academic with a PhD in energy engineering and predoctoral training at the Lawrence Berkeley National Laboratory (UC Berkeley)—appeared willing to rebuild relations with the academic community. She expressed the need for a new law to fund, promote, and regulate activities related to science, technology, the humanities, and innovation, along with the creation of a corresponding Ministry of Science, Technology, Humanities, and Innovation (Pérez Ortega and Basilio Reference Pérez Ortega and Basilio2024). Elevating CONACYT to the rank of a ministry could have given the agency greater political weight. Putting science, technology, the humanities, and innovation at the center of Mexico’s development strategy was also a positive sign. Yet, at the time of this writing, no bill to administer the new ministry has materialized, and allocated funds for science, technology, the humanities, and innovation have been reduced—both in real and percentage terms—by 2%, amounting to a cut of 1.3 billion pesos. Overall funding for Mexico’s 24 government-funded research centers dropped by 11%. These cuts are forcing some centers to reduce researcher salaries or lay off workers. Others report that they do not currently have enough funding to operate for the full year.Footnote 21
The academic community was also encouraged by the appointment of Rosaura Ruiz Gutiérrez to lead the new ministry. Ruiz holds a PhD in biology, completed postdoctoral training at the University of California, Irvine, and is a member of the Mexican Academy of Sciences. Most recently, she served as Secretary of Education, Science, Technology, and Innovation of Mexico City under then-Mayor Sheinbaum. Although she has indicated that austerity measures will continue, she has also pledged to seek greater funding and rebuild international collaborations. However, at the time of this writing, none of these promises have materialized, and it remains unclear how they could be achieved given current budget cuts. We can only hope that Sheinbaum and Ruiz, as former academics themselves, will demonstrate the political will to protect academic freedom for the sake of knowledge and democracy in the country.
Supplementary material
To view supplementary material for this article, please visit http://doi.org/10.1017/S1537592725102077.
Acknowledgments
We are grateful to Guillem Riambau for his valuable feedback. We also thank participants in the September 2024 Working Papers Workshop at the Department of Politics and International Affairs at Wake Forest University, attendees of the Democratic Erosion and Resilience panel at the 81st Annual MPSA Conference, and the anonymous reviewers at Perspectives on Politics for their insightful suggestions. We dedicate this study to the collective behind the protest movement #YodefiendoalCIDE.
Data replication
Data replication sets are available in Harvard Dataverse at: https://doi.org/10.7910/DVN/QX8HUS
Conflicts of Interest
María Inclán is currently affiliated with CIDE, and Víctor Hernández-Huerta was formerly affiliated with it. These affiliations are disclosed in the interest of transparency; the views expressed in this article are based on independent research and do not represent the official positions of CIDE.
Open access
