Abstract
Global hunger is one of the most pressing issues facing us today. What explains cross-national and longitudinal variations in hunger? Current explanations of development and health outcomes more generally emphasize intranational or world economic processes. This analysis provides an alternative view, emphasizing the role of liberal and illiberal world culture in explaining cross-national and longitudinal variations in hunger. To test the proposed global institutional theory of hunger, I analyze an expansive country-level dataset using random effects (RE) Tobit panel regression models, 2001–2021. The results provide empirical support for the proposed theorization, suggesting that multiple international social structures independently shape hunger rates beyond the domestic and world economic processes privileged in the previous literature.
Social scientists have long sought to understand the impacts of world culture on domestic nation-states and societies (e.g. Hironaka, 2014; McKinney et al., 2024; Meyer et al., 1997; Nachtigal, 2025; Ramirez et al., 1997; Shandra et al., 2004, 2008; Tuncer-Ebetürk et al., 2025; Ninomiya n.d).I move these conversations and an emerging literature on polarization in world society (e.g. Bromley et al., 2020; Furuta et al., 2023; Furuta and Bromley, 2025; Hadler and Symons, 2018; Lerch et al., 2022, 2024; Song, 2025; Song et al., 2024; Rotem 2026; Mejia 2024a) in important new directions by theorizing how liberal and illiberal world culture independently have beneficial effects on country-level measures of hunger. My argument is twofold. In liberal world society, a human rights cultural model of hunger was institutionalized during major international conferences of the 2001-2021. This cultural script specifies that hunger is a human rights violation, and nation-states/international organizations bear responsibility for addressing this issue. Illiberal world society actors frame hunger as a symptom of “global capitalism” and the “neoliberal order,” provide benefits to membership that reduce hunger, and even align their efforts on hunger with liberal world cultural ideals. These respective cultural models of hunger then orient action for actors across world society, thereby generating an uneven drift toward hunger reduction (Hironaka, 2014; Schofer et al., 2012; Schofer and Hironaka, 2005; Mejia 2025a).
To test these ideas, I conduct statistical analyses of an expansive cross-national, longitudinal dataset. Results from random effects (RE) Tobit panel regression models provide consistent empirical support for the proposed theoretical arguments, highlighting the independent influence of liberal and illiberal world cultural processes alongside domestic and global economic factors in shaping hunger rates. This study speaks to a burgeoning interdisciplinary literature on the cross-national, longitudinal sources of hunger by calling attention to its global liberal and illiberal cultural foundations.
The empirical problem: Global trends in hunger
The international community has made tremendous progress in eradicating global hunger. Figure 1 illustrates this point by plotting global-level trends in hunger from 2001 to 2021, which is measured and interchangeably referred to as a country’s prevalence of undernourishment (PoU). The first decade of the 21st century saw dramatic declines in global hunger, reaching just under 11 percent in 2010 and generally plateauing for some time. However, over 700 million people still face hunger in the world, with many being women and children (Von Grebmer et al., 2023). Hunger also has dramatic implications for the development and health of those experiencing it, which may raise the possibility of early child deaths; future ailments such as cardiovascular disease, diabetes, and obesity; and falling into extreme poverty (Black et al., 2013; Galasso et al., 2016). Furthermore, undernourishment negatively impacts economic growth (McGovern et al., 2017). Even more striking is that, in recent years, global hunger rates began steadily increasing. With these trends in global hunger and their implications for society/vulnerable demographics, the question is obvious: What explains cross-national and longitudinal variations in hunger?

Global trends in hunger, 2001–2021.
Current explanations
Current explanations of development and health more generally point to economic development, fertility and education, women’s legislative representation, and government political structure. Others emphasize the role of economic globalization, where particular attention is paid to trade. I turn to an overview of these respective arguments accordingly. 1
Economic development
One view is that economic development decreases hunger. There are many arguments made in the literature. One is that economic development provides a greater supply of food for domestic consumption (Smith and Haddad, 2002). Second is that increases in affluence can lead to greater wages and standards of living more generally (Firebaugh and Beck, 1994), which may allow people to purchase and have more consistent access to food supplies (Lange and Vollmer, 2017). Third is economic development which may contribute to increased social services, education, and nutrition for domestic populations (Anand and Ravallion, 1993; Subramanyam et al., 2011). Mechanisms are still somewhat speculative and many point to limitations of this relationship (Brady et al., 2007; Page and Pande, 2018; Sen, 1999), but a long line of research spanning decades indeed suggests that economic development exerts beneficial effects on social welfare outcomes (Collier, 2008; Dollar and Kraay, 2004; Easterly, 1999; Lange and Vollmer, 2017; Preston, 1975; Pritchett and Summers, 1996; Ravallion, 2015; Sachs, 2005; VanHeuvelen and VanHeuvelen, 2019; Mejia 2022a). I thus test the conventional view that economic development is inversely associated with hunger.
Fertility
One argument made in the literature is that high fertility rates increase hunger as it further strains the budgets of already resource-scarce families, leading to more individuals experiencing hunger (Birdsall and Griffin, 1988). Another argument is that higher rates of fertility result in decreased wages and broader economic growth within a country, which would then increase hunger (Eastwood and Lipton, 1999). Third is that higher fertility rates may contribute to the generation of poverty traps. Studies find that fertility exerts harmful effects on a variety of health outcomes (e.g. Austin and Hof, 2023; Mejia 2023a, 2023b, 2024b). I thus expect that fertility is positively associated with hunger.
Education
The role of education is also emphasized in the literature. Numerous arguments are made. It could be that increased education is associated with increased wages as individuals accumulate more human capital (Caldwell, 1979; Hatt and Waters, 2006). Increased wages then help an individual purchase and maintain access to food, which then decreases hunger. Another argument is education may decrease fertility, which would then decrease the number of individuals that scarce resources need to be distributed to (Cleland and Van Ginneken, 1988; Gakidou et al., 2010; Mejia 2025c). Furthermore, increased education can lead to increased political mobilization, where individuals can demand for increased dietary and nutritional resources (Austin et al., 2014). These theoretical considerations suggest that secondary school enrollments is inversely associated with hunger.
Women’s legislative representation
Another prominent view is that increases in women’s legislative representation benefit health and social welfare outcomes. Arguments are quite diverse, but some points made are that greater gender parity in legislative representation is associated with increased spending on social services and decreased poverty (Bolzendahl, 2009; Bolzendahl and Brooks, 2007; Brady, 2009; Funk and Philips, 2019). Others argue that increases in women’s legislative representation lead to increases in women and children rights bills (Swiss et al., 2012). It is also argued that women’s legislative representation is associated with increased economic growth, which would then affect social welfare/health outcomes (Dahlum et al., 2022). Furthermore, it is posited that increases in women’s legislative representation is associated with increased resources available to children and women (Hornset and De Soysa, 2022; Wyant, 2021), which would then lead to decreased hunger, especially as women and children disproportionately represent those facing it (Von Grebmer et al., 2023). Numerous studies find empirical support for these theoretical arguments (Quamruzzaman and Lange, 2016; Rustagi and Akter, 2022; Swiss et al., 2012). I thus expect that women’s legislative representation is inversely associated with hunger.
Democracy
A country’s political structure is another potential explanation of cross-national and longitudinal variations in hunger. Many of the arguments made in the literature revolve around electoral dividends for political leaders. In more democratic countries, politicians may be pressured to provide social assistance/government aid to garner political support for election and remain in office once elected (Frey and Al-Roumi, 1999; Lake and Baum, 2001), which would then decrease hunger. With increased civil and political liberties, constituents may be able to pressure governments to provide social assistance and legislators may face negative political consequences if they do not follow-up on their campaign promises or fail to provide such services when in office (Ruger, 2005). Autocracies, however, face different electoral dynamics. Autocracies may be less responsive to the needs of their constituents relative to democracies, where their political leaders may be more interested in portraying a positive impression to others rather than providing government assistance to their constituents (Besley and Kudamatsu, 2006; Ruger, 2005). Autocracies may also suppress demands from their constituents to provide social services and government aid (Brown and Hunter, 1999). 2 I thus assess the hypothesis that democracy is associated with decreased hunger in the ensuing analyses.
Economic globalization
There is a theoretical tension in the literature over how aspects of economic globalization affect hunger. 3 This theoretical tension is generally composed of neoclassical economic theory and world-systems analysis/dependency theory. Neoclassical economic theory suggests that trade benefits hunger. Regarding international trade, classic work suggests that international trade increases economic growth as countries pursue their “comparative advantage” in the global economy (Magee, 1980) or through the transfer of technologies across borders (Jenkins and Scanlan, 2001; Mejia 2025b). This economic growth then benefits hunger. Trade is also argued to increase access to food and nutritional diversity beyond what can be produced in an individual country (Mary, 2019). Extending this logic, greater integration into global economic trade markets may allow for increased food security for a country’s population, which would decrease hunger over time (Dithmer and Abdulai, 2017). World-systems analysis/dependency theorists argue that trade exerts harmful effects on hunger. Arguments in the literature are effectively the opposite of what neoclassical economic theory would predict, where it is particularly argued that shifts to an export-oriented economy divert land from “food production to commercial purposes, reducing the amount of food available for domestic consumption” (Jenkins and Scanlan, 2001; Shen and Williamson, 1997: 670). Others argue that increased reliance on trade may make a host country “more vulnerable to international market forces and allow [countries in more advantageous positions in the world economy] to obtain favorable terms of trade” (Kentor and Boswell, 2003: 303), which might increase hunger rates over time. 4 In summary, neoclassical economic theory argues that trade is inversely associated with hunger, while world-systems analysis/dependency theory posit that trade is positively associated with hunger.
The argument: A global institutional theory of hunger
Here, I present an alternative theory to explain cross-national and longitudinal variations in hunger. I begin by delving into the central tenets of world society theory and an overview of previous research. I then delve into the theorization of how liberal world culture shapes hunger. Next, I propose arguments regarding how linkages to illiberal world society also benefit hunger.
Liberal world society
The world society theory is a prominent paradigm in the comparative international social sciences. This theory calls attention to a global social structure and concomitant culture that affects nation-states, domestic societies, and even individuals (Frank and Meyer, 2002; Meyer et al., 1997). This culture becomes embedded in organizations (especially those operating at the international level) and we can make inferences about these codified cultural frames by analyzing organizational structures (Hironaka, 2014; Inoue and Drori, 2006).
Much of the world society literature has helped advance theoretical and empirical understanding of global isomorphic trends. For example, numerous studies analyze environmental legislation/organizational structures (i.e. Frank, 1999; Frank et al., 2000; Hironaka, 2002, 2014; Longhofer et al., 2016), while others analyze issues such as the death penalty (Mathias, 2013), sex (Frank et al., 2010; Frank and McEneaney, 1999), population control (Robinson, 2015), abortion (Boyle et al., 2015), foreign aid (Swiss, 2016), and human rights (Wotipka and Tsutsui, 2008). A more specific body of literature draws on insights from the world society theory to explain cross-national, longitudinal variations in development outcomes (Kerrissey and Schuhrke, 2016; Shandra et al., 2010).
Building on this literature, I present generalizable, theoretical arguments about how liberal world society shapes domestic nation-state hunger rates. I argue that human rights cultural models of hunger were institutionalized in liberal world society toward the end of the 20th century, where the existence of hunger is viewed as a human rights violation and the responsibility of governments/international organizations to eradicate it. This cultural frame of hunger then helps mobilize global efforts to end hunger, which then has empirical implications for domestic nation-state hunger rates.
I illustrate these tenets by providing a historical overview of hunger in world society. The objective experience of hunger has effectively been a social issue ab immemorabili, but global efforts to eradicate it are scant, except in the recent decades (for a brief historical overview, see Ellison, 2004). The temporal scope of the statistical analyses is from 2001 to 2021, but uncovering the historical institutionalization of hunger in world society requires a broader purview. Pre-World War Two, hunger was generally understood as a matter that humanitarian and faith-based organizations addressed rather than the responsibility of nation-states and international organizations (Chabbott, 2015; Ferris, 2005; Houlihan, 2024). There was various international activity on the issue of food and nutrition such as the creation of the Food and Agriculture Organization (FAO) in 1945 following the 1943 UN Conference on Food and Agriculture. Also, there was the establishment of the World Food Programme in 1961, the mentioning of the “right of everyone to be free from hunger” in the International Covenant on Economic, Social, and Cultural Rights (ICESCR), and other international activities (the 1974 First World Food Conference; World Food Day established in 1981; 1992 First International Conference on Nutrition). However, human rights cultural models of hunger did not emerge until the 1996 World Food Summit in Rome, Italy. A major outcome of this event was the Rome Declaration on World Food Security, which writes that “We, the Heads of State and Government . . . reaffirm the right of everyone to have access to safe and nutritious food, consistent with the right to adequate food and the fundamental right to everyone to be free from hunger” (FAO, 1996, italics added). This declaration also states that We pledge our political will and our common and national commitment to achieving food security for all and to an ongoing effort to eradicate hunger in all countries, with an immediate view to reducing the number of undernourished people to half their present level no later than 2015. (FAO, 1996)
I posit that the World Food Summit institutionalized human rights cultural models of hunger in liberal world society.
These world cultural human rights models of hunger then diffuse to nation-states, domestic societies, and individuals. One mechanism is through international nongovernmental organizations (INGOs; Boli and Thomas, 1997; Lerch, 2019), where it is argued that country-level memberships in INGOs serve as a proxy of world society’s overall effect on nation-states (see Hironaka, 2014). Another proxy of world society embeddedness is a country’s ratifications of international treaties, where such ratifications can signal commitment to liberal world cultural principles (Frank et al., 2000). More general indexes are also used to capture the broader impact of world society (e.g. Schofer and Hironaka, 2005; Wang and Schofer, 2018).
Cultural frames codified in these models orient action on the issue of hunger. I posit that such dynamics have empirical implications for actual domestic nation-state hunger rates. Cultural frames are at the core of how world society generates change as they orient action for actors across world society (Hironaka, 2014; Schofer et al., 2012; Schofer and Hironaka, 2005). For example, if hunger is viewed as an unfortunate aspect of the human condition and the responsibility of charity/religious organizations, then global efforts to eradicate hunger are unlikely to arise. However, if hunger is understood as a human rights violation and human rights more generally are already valued in a society, then efforts to eliminate it might emerge across multiple levels of world society (Boyle et al., 2002; Frank et al., 2000; Swindle, 2023). Thus, numerous external pressures bear on nation-states/international organizations to abide by world cultural scripts of hunger, which creates an uneven drift toward actual improvements over time (Hironaka, 2014; Schofer and Hironaka, 2005). Per the above theorization, I argue that linkages to liberal world society is inversely associated with hunger.
Illiberal world society
A new body of world society literature suggests that illiberal cultural models also propagate in world society more generally. These studies find that illiberal world culture has harmful effects on a variety of important outcomes such as academic freedom, education enrollments, measures of women’s rights and gender equality, and civil society (Bromley et al., 2020; Lerch et al., 2022, 2024; Schofer et al., 2022; Velasco, 2023).
I argue that illiberal world culture has a beneficial effect on hunger. There are numerous potential mechanisms underlying the hypothesized inverse statistical relationship between illiberal intergovernmental organization (IGO) linkages and hunger. I illustrate that, rather than backlashing toward liberal world cultural principles regarding the right to be free from hunger per se, illiberal world society actors frame the issue of hunger as an ill of “global capitalism” and the “neoliberal order” or in similar ways as liberal world culture. Illiberal IGOs also provide tangible, material benefits to member-states that help reduce hunger.
Illiberal world society actors propagate frames that view hunger as a function of global capitalism and the “neoliberal order.” For example, at the 1996 World Food Summit, Fidel Castro lamented that the UN plan to reduce world hunger by 50 percent over the next two decades as a “shamefully modest objective” and that “Capitalism, neoliberalism, the laws of a wild free trade marketplace, external debt, underdevelopment . . . kill people” (Agnew, 1996). Castro also commented that “Hunger, the inseparable companion of the poor, is the offspring of unequal distribution of wealth and injustices in this world. The rich do not know hunger” (Agnew, 1996).
Similar discourses on hunger continued at later events. At an emergency 2008 meeting of the ALBA-TCP member nations, Venezuelan President Hugo Chavez commented that the world food crisis is “the greatest demonstration of the historic failure of the capitalist model,” while Cuban Vice President Carlos Lage stated that it is part of an “unjust international economic order . . . [in which] the logic is profit and not the satisfaction of people’s needs.” Bolivian President Evo Morales later commented that “it is the responsibility of presidents to act in concert to guarantee the food security of our peoples” (Suggett, 2008). ALBA-TCP member nations later signed the Agreement for the Implementation of Cooperation Programs for Food Security and Sovereignty. This agreement helped lead to a variety of initiatives to combat hunger such as food security funds, food distribution networks that cater to each member country’s production chains, and funding agribusiness development to increase food production among ALBA-TCP member nations. ALBA-TCP member countries, lamenting the imposition of an “unequal” and “unjust” neoliberal economic order, signed another declaration in a later presidential summit that rejects subsidies from “unfair” trade with developed countries and proposed an action plan to strengthen member countries’ economies and food production systems (Joubert-Ceci, 2008). Similar sentiments are also evident in the Agreement on Food Security and Sovereignty of the Member Countries of PETROCARIBE and ALBA, which writes that “food sovereignty is defined as the rights of peoples, their countries or unions of states to define food policy towards third countries, leaving aside unfair international trade policies . . .” (FAO, 2024). Among various plans of action, a goal of this Food Security and Sovereignty Agreement is “development of actions . . . that lead to increasing food production . . . for the national self-sufficiency of each party. If there are surpluses, they will be distributed through distribution and exchange mechanisms between the member countries” (FAO, 2024). Furthermore, at the 2013 Second Summit of ALBA and Petrocaribe, attendees agreed to implement an Action Plan for [the] Eradication of Hunger and Poverty (Pearson, 2013) among member countries.
Illiberal world cultural values can sometimes align with liberal world society. For example, in 2017, the Economic and Social Council of the League of Arab States approved the establishment of an Arab High Level Sustainable Development Committee to help coordinate efforts to reach the Sustainable Development Goal (SDG) of “Zero Hunger” in the region. In a later meeting, a Subcommittee for Hunger Eradication, Food Security, and Sustainable Agriculture in the Arab Region was established, where the goal of this committee was to monitor progress toward reaching the second SDG of eradicating world hunger (FAO, 2022). This committee is also tasked with developing an action plan specific to the Arab region to reach the SDGs (FAO, 2022). The League of Arab States has even organized the 2018 Arab Sustainable Development Week, where the conference aimed at strengthening partnerships among countries to achieve the SDGs, with the second SDG being eradicating hunger in the world by 2030. Similarly, the Gulf Cooperation Council (GCC) endorsed global efforts to eradicate world hunger and poverty and distributes aid to member states and non-GCC countries (Saudi Press Agency, 2023). Another example is the Shanghai Cooperation Organization (SCO), which also seeks to work toward reaching the SDG goal of “Zero Hunger” (CGTN, 2023).
The focus thus far has been on cultural values codified in international organizations, but the concrete actions of illiberal organizations can also affect hunger. Of relevance here is the potential benefits allocated to nation-states per their membership in illiberal IGOs (Cottiero and Haggard, 2023). For example, the SCO is prescribed with responsibility for improving access to physical capital to increase agricultural production, economic growth, development, and the living standards of people more generally in the SCO region (Shanghai Cooperation Organization, 2020). Similarly, ALBA-TCP, after lamenting that hunger is linked to the “oligopolistic nature of the large transnational corporations that control the food agricultural industry,” posited that ALBA-TCP members will provide joint investments and “accelerate food production . . . [and] distribution in the region” (Granma, 2008).
There are additional reasons to expect that affiliation with illiberal IGOs may be associated with lower levels of hunger. For example, linkages to illiberal world society may expand access to large non-Western markets such as China for export goods, which can support economic growth in exporting countries (Assadi, 2025). Such linkages may also help secure trade routes and diversify food imports, potentially reducing vulnerability to disruptions in global supply chains.
In other cases, illiberal IGOs establish specialized coordinating institutions focused explicitly on agriculture and food security. For instance, the Arab League created the Arab Organization for Agricultural Development (AOAD) and the Crisis Management Department (CMD). The former is responsible for coordinating agricultural policies and food security in member-states, while the latter partners with the United Nations Development Programme (UNDP) to strengthen crisis-response strategies for its member-states. Through initiatives such as the Arab Program for Sustainable Food Security, the AOAD has promoted crop production, water management, agricultural research infrastructure, and regional economic integration aimed at stabilizing long-term access to food supplies.
Overall, illiberal IGOs may reduce hunger through a variety of pathways. This includes oppositional framing and selective alignment with liberal world culture. Furthermore, illiberal IGOs may provide material resources to affiliated countries that can reduce hunger. I thus also argue that linkages to illiberal world society is associated with decreased hunger.
Data and methods
Sample
I analyze a global sample of over 130 countries from 2001 to 2021 with close to 2500 observations; 908 observations are left-censored, while around 1500 observations are uncensored. I provide descriptive statistics and correlation matrix in Appendix 1, while Appendix 2 provides a list of the countries analyzed.
Dependent variable
Hunger
The response variable of interest is a country’s PoU, which captures the percentage of the population whose habitual food consumption is insufficient to provide for the dietary energy levels that are required to maintain a normal active and healthy life. These data are from the FAO and provided by the World Bank’s (2023) World Development Indicators.
Independent variables of interest
Liberal world society linkages
The world society literature has long emphasized that world cultural principles become embedded in organizations and diffuse to domestic nation-states via organizations operating at the international level (Boli and Thomas, 1997; Hironaka, 2014; Lerch, 2019; Shandra, 2007). Of particular interest in the world society literature is country-level counts of memberships in INGOs. I include three measures of a country’s linkages to liberal world society. The first is a country’s total count of all INGOs, while the second is women’s rights INGOs (WINGOs). The third liberal world society proxy is the total count of INGOs that focus specifically on development issues (DINGOs). Each measure is logged to account for skew. The data for all INGO counts is from the Union of International Organizations Yearbook of International Associations (Union of International Organizations 2001–2021), while the WINGO and DINGO data is from the Transnational Social Movement Organization Dataset (Smith and Wiest, n.d.). I use linear interpolation to fill missing data and extrapolation to extend the data from 2013 to 2021. Carrying forward the 2013 values through to 2021 and extending the data using growth rates produced similar results and are available upon request. 5
Human rights treaty ratification
State-centric treaty measures proxies of liberal world cultural embeddedness are also emphasized in the world society literature (Boyle and Kim, 2009; Nyseth Brehm and Boyle, 2018) and in analyses of development outcomes (Palmer et al., 2009; Swiss et al., 2012). I thus include a count measure (ranging from 0 to 9) of whether a country has ratified the following human rights treaties: the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD), the International Covenant on Civil and Political Rights (ICCPR), the International Covenant on Economic, Social, and Cultural Rights, the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW), the Convention Against Torture or other Cruel, Inhuman, or Degrading Treatment or Punishment (CAT), the Convention on the Rights of the Child (CRC), the International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families (ICMW), the International Convention for the Protection of All Persons from Enforced Disappearance (CPED), and the Convention on the Rights of People with Disabilities (CRPD). Hunger is framed as a human rights matter in liberal world society, making it possible that increases in a country’s number of human right treaty ratifications more generally captures their commitment to liberal world cultural principles and is thus associated with decreased hunger.
Liberal world society index
World society research also incorporates broader index measures of liberal world society effects (Lerch et al., 2024; Wang and Schofer, 2018). I thus incorporate a liberal world society index by summing the z-scores of the above measures.
Illiberal world society linkages
Following Lerch et al. (2024), I capture the effects of illiberal world society on domestic nation-states by including a measure of country-level memberships or observer-/partner-type status in the following IGOs: Alianza Bolivariana para los Pueblos de Nuestra América (ALBA), Commonwealth of Independent States (CIS), GCC, League of Arab States (LAS), and the SCO. This proxy builds off the indices used in Bromley et al. (2020), Lerch et al. (2022) and Schofer et al. (2022) (see Lerch et al., 2024 for an in-depth description of these organizations).
Control variables
Economic development
The analyses control for the natural logarithm of real GDP per capita in purchasing power parity (PPP). These data are from the World Bank’s (2023) World Development Indicators.
Fertility
I incorporate measures of a country’s total fertility rate, which is the number of children that would be born to a woman if she were to live to the end of her childbearing years and bear children in accordance with age-specific fertility rates of this specific year. These data are from the World Bank’s (2023) World Development Indicators. 6
Education
The analyses include secondary school enrollments as a percentage of gross. These data are gathered from the World Bank’s (2023) World Development Indicators.
Women’s legislative representation
I include a measure of women’s legislative representation from the V-DEM dataset (Coppedge et al., 2023), which captures the percentage of a country’s lower or unicameral of the legislature that are female.
Democracy
I also incorporate measures of a country’s political structure using the Polity5 Project dataset, where −10 is strongly autocratic and 10 is strongly democratic (Marshall and Gurr, 2020). 7
Economic globalization
I include a common proxy of economic globalization: trade openness. This measure is operationalized as trade (exports + imports) as a percentage of GDP. I take the natural logarithm of these covariates to address skew. The data on trade are from the World Bank’s (2023) World Development Indicators. 8
Methods
To analyze the cross-national longitudinal data, I incorporate RE Tobit panel estimation. RE Tobit models are a commonly used econometric method to analyze panel data when the response variable is censored. 9 Censoring refers to when the value of a variable is known only up to a certain value. The PoU data are censored at a lower limit of 2.5. Using traditional linear regression models on censored data—whether analyzing all observations (censored and uncensored) or only uncensored observations—produces inconsistent parameter estimates (Wooldridge, 2013: 610). 10 RE Tobit panel estimation addresses this problem by explicitly accounting for the censoring mechanism through maximum likelihood estimation (MLE), producing much more reliable parameter estimates relative to incorporating traditional linear panel estimation methods for censored data (Wooldridge, 2013: 610). The statistical notation of RE Tobit panel regression models is as follows
where
where
Results
Table 1 reports the analysis results. Beginning with the covariates of interest, I observe that counts of country-level memberships in all INGOs are negative in direction and statistically significant. 11 Human rights treaties do not reach statistical significance, but human rights INGOs and development INGOs are independently negative in direction and significant. The liberal world society index is also negative in direction and significant. Turning to the measure of illiberal IGO linkages, I observe that the unstandardized coefficients reach statistical significance and are negative in direction. Apart from the human rights treaties measure, I consistently find empirical evidence for the proposed theorization. Liberal and illiberal world cultural processes independently shape global hunger rates.
Random effects Tobit panel regression models of hunger, 2001–2021.
Two-tailed tests; unstandardized coefficients flagged for significance; standard errors in parentheses.
p < .1; *p < .05; **p < .01; ***p < .001.
I now turn to discussing the results from statistical controls. I observe that GDP per capita is consistently significant and negative in direction throughout the reported models of Table 1. Fertility is surprisingly negative in direction and significant in the models reported throughout Table 1. I observe that secondary school enrollments (% of gross) are consistently negative in direction and significant. Similarly, increases in women’s legislative representation are associated with decreased rates of hunger in Table 1. Democracy, however, does not consistently reach statistical significance. Trade openness is significantly associated with an increase in hunger, which aligns with previous work on the harmful effects of world economic integration (Mejia 2021, 2022b, 2024c). Figures 2 and 3 provide predictive margins with shaded 95 percent CIs, while Figure 4 plots hunger rates by illiberal IGO membership.

Predictive margins for liberal world society measures with shaded 95 percent confident intervals.

Predictive margins for illiberal IGOs with shaded 95 percent confident intervals.

Hunger rates by illiberal IGO membership, 2001–2021.
Discussion and conclusion
Social scientists have long sought to explain cross-national and longitudinal variations in hunger. This is important because over 700 million people, largely women and children, still experience hunger. Also, hunger rates have been increasing in recent years. We need social scientific understanding of the socio-structural sources of hunger.
Much dialogue focuses on the role of economic development (Collier, 2008; Lange and Vollmer, 2017; Pritchett and Summers, 1996; Soriano and Garrido, 2016), fertility and education (Caldwell, 1979), women’s legislative representation (Rustagi and Akter, 2022; Swiss et al., 2012), and government political structure (Salazar et al., 2023). Others debate the effects of trade (Austin, 2012; Austin et al., 2012; Mary, 2019).
In this analysis, I propose a new theoretical approach to explain cross-national and longitudinal variations in hunger. Extending the world society theory (Meyer et al., 1997) and the more specific literature on health and development (Brass et al., 2018; Inoue and Drori, 2006; Shandra et al., 2010), I theorize how liberal and illiberal world culture independently affect hunger rates. A human rights cultural model of hunger was institutionalized in liberal world society during major international conferences in the 2001-2021, specifying that hunger is a human rights violation and the responsibility of nation-states/international organizations to reduce. In illiberal world society, hunger is culturally viewed as a by-product of global capitalism’s “tentacles” and neoliberal economic policy, in a similar manner as liberal world culture, or as an issue illiberal IGOs must alleviate for their members. I contend that these global cultural processes have dramatic implications for hunger.
To test my theoretical arguments, I estimate RE Tobit panel regression models of country-level, cross-sectional, longitudinal data from 2001 to 2021. The statistical analyses suggest that, as the respective liberal and illiberal world society linkages measures increase, hunger declines, ceteris paribus. These results are also robust to a variety of alternative econometric specifications.
There are numerous theoretical contributions that stem from the analysis. First is that I develop a macro-level, neo-institutional theory of hunger and illustrate its remarkably robust explanatory power using powerful econometric methods to analyze expansive panel data on over 130 countries. Future research seeking to explain cross-national, longitudinal variations in hunger must not only account for the “bottom-up” and global economic theoretical perspectives offered in the previous literature, but the global cultural and social environment that nation-states are embedded in—world society. These findings do not diminish the importance of domestic and global economic theoretical explanations, but rather reveal that global institutional processes also shape hunger alongside many established predictors of social development. Furthermore, the statistical analyses suggest that, for countries that are in the lower percentile in terms of linkages to liberal (low count of INGOs) and illiberal (no affiliations with illiberal IGOs) world society linkages, their hunger levels see some decline as these respective world society proxies increase, ceteris paribus. I also observe that beneficial effects to liberal world society arise from WINGOs, DINGOs, and the general INGO measure, whereas previous research emphasizes the importance of issue-specific INGOs (i.e. Velasco, 2018; Mejia 2020). These specific findings deserve more attention in future work, but it could be that various social processes arising from the activities of issue-specific INGO efforts (i.e. a DINGO providing food aid) and INGOs (i.e. linkages liberal world cultural models of hunger) more generally independently affect hunger rates.
The second theoretical contribution of the analysis is that I explore the boundaries of world society theory and social welfare. This larger contribution is composed of two components. First, I extend the existing literature (Brass et al., 2018; Chabbott, 2015; Inoue and Drori, 2006) by illustrating that world society affects a broader set of outcomes than the existing literature would predict (Noble and Austin, 2014; Shandra et al., 2010). Second, I theorize how illiberal world culture affects social welfare and development, where previous literature focuses on the impacts of liberal world culture (Kerrissey and Schuhrke, 2016; Noble and Austin, 2014; Shandra et al., 2010). Third, I show that illiberal world society actors do not just react to expanding liberalism (Jepperson and Meyer, 2021), but propagate their own cultural scripts.
This study also moves an emerging literature on illiberal world culture in important new directions. Recent world society studies call attention to an emerging illiberal backlash to expanding liberalism and document its empirical implications (e.g. Bromley et al., 2020; Lerch et al., 2022, 2024; Schofer et al., 2022; Velasco, 2023). Here, in contrast to this literature, I observe that linkages to illiberal world culture have beneficial effects on hunger and theorize how this may possibly occur. Of course, this is not to say that illiberal international structures are always beneficial. Rather, it appears that hunger is a case of homogeneous liberal and illiberal world cultural effects.
These findings generate important directions for future research. We need additional research on the impacts of liberal and illiberal world culture. As Jepperson and Meyer (2021: 310) note, a key dynamic of the post-liberal era in world society is “the conflict between expanding liberalism and its critics.” It appears that illiberalism is gaining legitimacy in world society and thus shaping global change in ways that would be difficult to imagine in earlier world society eras (Bromley et al., 2020; Lerch et al., 2022, 2024; Schofer et al., 2022; Velasco, 2023). We thus need continued analyses of liberal and illiberal world culture. Such analyses may advance our understanding of which cultural models are becoming institutionalized in world society and their empirical effects.
There are some limitations to the analysis that warrant attention. First, the statistical analyses cannot definitively speak to the determinants of the post-2017 aggregate-level global increases in hunger observed in Figure 1. The observation that trade openness is positively associated with hunger gives insight into some possibilities, but it is important to note that the results reported here can only safely allow us to infer the factors explaining cross-national, longitudinal variation in hunger rather than specific aspects of the aggregate-level trends in Figure 1. 12 Future research should incorporate research designs that may be able to attend to these questions in a scientifically sound manner. Second is the potential for measurement bias, where it is possible that countries with linkages to illiberal world society may underreport their hunger rates (Pandian, 2025). It is difficult to definitively assess whether such processes are occurring in the generation of the national-level PoU data provided by the FAO, but this is an important issue for future research to explore if such concerns become apparent.
To summarize, this analysis provides a new theoretical framework to explain cross-national and longitudinal variations in hunger, where I theorize the independent effects of liberal and illiberal world culture. In doing so, I illustrate that understanding the social–structural determinants of country-level hunger rates requires not only an accounting of “bottom-up” and global economic processes, but of the global social and cultural environment. These findings suggest that competing international structures penetrate domestic societies, helping their populations secure access to the nutritional foundations of sound individual, health, growth, and development.
Footnotes
Appendix 1
Descriptive statistics and correlation matrix.
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) Prevalence of undernourishment | 1.000 | ||||||||||||
| (2) GDP per capita (ln) | −0.732 | 1.000 | |||||||||||
| (3) Fertility rate | 0.702 | −0.795 | 1.000 | ||||||||||
| (4) Secondary school enrollments (% gross) | −0.711 | 0.844 | −0.841 | 1.000 | |||||||||
| (5) Women legislators (%) | −0.173 | 0.264 | −0.206 | 0.320 | 1.000 | ||||||||
| (6) Democracy | −0.268 | 0.385 | −0.375 | 0.398 | 0.206 | 1.000 | |||||||
| (7) Trade openness (ln) | −0.124 | 0.304 | −0.289 | 0.246 | 0.036 | 0.065 | 1.000 | ||||||
| (8) INGO linkages (ln)_i | −0.626 | 0.750 | −0.643 | 0.676 | 0.349 | 0.487 | −0.032 | 1.000 | |||||
| (9) HR Treaties | −0.194 | 0.021 | −0.111 | 0.166 | 0.301 | 0.216 | −0.029 | 0.138 | 1.000 | ||||
| (10) WINGO linkages (ln) | −0.409 | 0.453 | −0.368 | 0.393 | 0.272 | 0.527 | −0.180 | 0.856 | 0.142 | 1.000 | |||
| (11) DINGO linkages (ln) | −0.107 | 0.124 | −0.037 | 0.049 | 0.228 | 0.408 | −0.273 | 0.595 | 0.099 | 0.822 | 1.000 | ||
| (12) Liberal world society index | −0.449 | 0.471 | −0.399 | 0.436 | 0.364 | 0.540 | −0.167 | 0.877 | 0.376 | 0.950 | 0.835 | 1.000 | |
| (13) Illiberal IGO linkages | −0.045 | −0.025 | −0.060 | 0.027 | −0.219 | −0.501 | 0.021 | −0.204 | −0.077 | −0.373 | −0.505 | −0.379 | 1.000 |
Appendix 2
Acknowledgements
The author thanks the anonymous reviewers for helpful suggestions to improve previous drafts of the manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability statement
All data used to produce the results presented in the manuscript will be shared upon reasonable request to the corresponding author.
