https://orcid.org/0000-0001-6549-4687
Abstract
How has the composition of domestic groups testifying before Congress on foreign aid changed over time? In what ways have the main themes of their speeches changed? At US congressional hearings, numerous external groups have expressed their views and positions on aid issues, with some repeatedly invited to testify to multiple committees across many years. However, there is a lack of data on this representation to show how domestic politics have influenced donor's aid decision-making process. I present new findings on the evolution of the aid topics discussed by witnesses representing NGOs, private enterprises, and research experts based on their combined total of 1,656 testimonies given to 31 committees at 325 relevant hearings from 1980 to 2020. This paper's findings shed light on the various aid demands made by different group types, offering insights into the dynamics of aid discussions in the United States over the past four decades.
¿Cómo ha cambiado con el tiempo la composición de los grupos nacionales que testifican ante el Congreso sobre la ayuda internacional? ¿De qué maneras han cambiado los temas principales de sus discursos? En las audiencias del Congreso de los Estados Unidos, numerosos grupos externos han expresado sus puntos de vista y sus posiciones en materia de cuestiones relativas a la ayuda, y algunos de estos grupos han sido invitados repetidamente a testificar ante múltiples comités a lo largo de muchos años. Sin embargo, existe una falta de datos sobre esta representación que muestren cómo la política interna ha influido sobre el proceso de toma de decisiones con respecto a la ayuda por parte de los donantes. Presentamos nuevas conclusiones relativas a la evolución de los temas de ayuda que fueron debatidos por los testigos, quienes representan a ONG, empresas privadas y expertos en investigación. Para ello, partimos de la base de un total combinado de 1656 testimonios prestados ante 31 comités en 325 audiencias relevantes desde 1980 hasta 2020. Las conclusiones de este artículo arrojan luz sobre las diversas demandas en materia de ayuda realizadas por diferentes tipos de grupos, ofreciendo, de esta manera, una visión de la dinámica de los debates en materia de ayuda en los Estados Unidos durante las últimas cuatro décadas.
Comment la composition des groupes nationaux témoignant au Congrès américain concernant l'aide étrangère a-t-elle évolué ? Quelles sont les modifications de thèmes principaux dans leurs discours ? Lors des auditions du Congrès des États-Unis, nombre de groupes externes ont exprimé leur point de vue et leur position quant aux questions d'aides, certains ayant été invités plusieurs fois à témoigner devant nombre de comités des années durant. Néanmoins, il manque des données s'agissant de cette représentation pour montrer comment les politiques nationales ont influencé le processus de prise de décisions relatif aux aides du donateur. Je présente de nouvelles conclusions sur l’évolution des sujets d'aides abordés par des témoins représentant des ONG, des entreprises privées et des experts chercheurs se fondant sur un total combiné de 1 656 témoignages offerts à 31 comités lors de 325 auditions pertinentes entre 1980 et 2020. Les conclusions de cet article mettent en lumière les diverses demandes d'aides émises par différents types de groupes, ce qui nous renseigne sur les dynamiques des discussions sur les aides aux États-Unis ces quarante dernières années.
Introduction
How do domestic groups influence politicians to reflect their preferences in the foreign aid policymaking process? In the literature on legislative politics, congressional hearings in which these external groups lobby legislators on given policies have been studied exhaustively (Austen-Smith 1993; Grossman 2005; Schnakenberg 2017). For legislators, in the face of limited time, information acquisition is viewed as an essential need in making decisions about legislation (Krehbiel 1992). Studies note that by inviting witnesses to testify at hearings, legislators not only acquire information about an issue but also signal their support for an idea by using the information to promote their views and preferences (Hay 1997; Diermeier and Feddersen 2000; Brasher 2006; McGrath 2013). Further, several scholars exploring the question of which interest groups and lobbyist groups are most likely to testify note that such groups are strongly motivated to appear at committee hearings (Leyden 1995; Kollman 1997; Holyoke 2003, 2008) because testifying before Congress is an effective way to advocate for their agenda in the decision-making process (Schlozman and Tierney 1986).
The theoretical importance of congressional hearings through which both politicians and domestic groups have an incentive to interact with one another is relatively well addressed in the literature. Yet, studies to date offer very little empirical data pertinent to establishing which groups present most frequently at these hearings under what political conditions or the topics that these groups emphasize. Several recent studies have shown how the policy preferences of the general public and interest groups are expressed and reflected in the legislative process using new data collected by the respective authors based on political actors’ speeches and tweets (Barberá et al. 2019; Maher et al. 2020; Ban et al. 2023).
By comparison, whether theoretically or empirically, much less attention has been paid to the connection between politicians and domestic groups in relation to foreign aid decision-making. In particular, researchers have rarely engaged with the critical question of how groups such as private corporations, non-governmental organizations (NGOs), and think tanks gain access to and come to influence legislators through providing testimony and information at committee hearings and so affect aid policymaking. There is a burgeoning literature exploring the effects of donors’ domestic politics on aid (Thérien and Noel 2000; Tingley 2010; Milner and Tingley 2011; Paxton and Knack 2012; Brech and Potrafke 2014; Heinrich et al. 2016; Dietrich et al. 2020). However, more attention has been given to how different worldviews, values, and beliefs shared within donor countries shape the overall direction of the states’ aid-giving.
Relatively few studies examine the impact of mobilized domestic interests in aid with a focus on interest group activity, although some recent empirical work does investigate the role of a broadly defined body of interest groups in aid decision-making (Roberts 2014; Milner and Tingley 2015; Dietrich 2021). The distributional consequences of aid are considered in this strand of research, which sheds light not only on the normative and moral values that domestic groups promote but also on the material needs they seek to meet in this process. For example, Milner and Tingley (2015) showed that interest groups push for budgets for economic aid through the legislative process by, for example, lobbying and testifying before Congress in order to win contracts. According to Roberts (2014), development contractors, including private enterprises and NGOs, influence the US aid decision-making process, especially in the case of the for-profit sector stressing the need for more contract awards.
Building on this work, my contribution to the literature centers on exploring witnesses’ speeches at US aid hearings as a way to consider how domestic groups attempt to influence the state's aid decisions. Which domestic groups have appeared at US congressional hearings for foreign aid? What kinds of aid issues have these groups focused on? Has the composition of the witness groups testifying before Congress on foreign aid changed over time? Have the main themes of their speeches changed? How do exogenous events and global norms and values affect which witnesses are invited to testify and their discussions about aid before Congress? Exploring these questions in empirical terms is important to furthering understanding of who congressional constituencies for aid are and what they want, which will, in turn, advance research efforts focused on identifying the domestic mechanisms through which donors make aid decisions.
In this paper, I present a computational analysis of testimonial statements given by non- governmental witness groups whose organizational goals and activities are directly related to US aid. These include private voluntary organizations, private enterprises, research-oriented experts, and think tanks focused on international aid and development, and relevant sectors such as education, health, and agriculture. I classify the witness organizations into three group types, i.e., NGOs, private enterprises, and research experts, as I expect to find differences among them in regard to the focal topics addressed in their testimonies and the likelihood of being (re)invited across different committee hearings.
In the next section, I first provide descriptive statistics on my data showing variations in terms of witness composition across years and committees. I then present quantitative evidence for the evolution of topic representations over time as well as according to witness group type. I acknowledge that my findings cannot speak to how external groups interested in aid engage with their governments’ decision-making processes beyond the United States. However, the findings do provide a basis for future work on how the distributional politics of aid work inside a donor country.
Data Collection
I collected data on testimonial statements given by non-governmental witnesses at congressional hearings in the period of 1980–2020 from ProQuest's congressional database using the search term foreign aid. As foreign aid is an autogenerated category in the database, which includes a variety of terms referring to foreign aid and its sub-categories such as development assistance, foreign assistance, economic aid, military aid, and food aid. A total of 2,073 records were returned for published hearings in ProQuest's congressional database searched with the term foreign aid from 1980 to 2020. Of these, any with a hearing title encompassing issues broader than foreign aid (e.g., US–Pakistan relations and foreign economic relations with Latin America), only the panel(s) and witness(es) whose testimonies focused principally on addressing foreign aid were included in the analysis. As stated earlier, my data comprise testimonies given by non-governmental witnesses whose organizational goals and activities are directly related to the US foreign aid industry, filtered by the witnesses’ affiliations. Testimonies in which witnesses explicitly indicated that they were appearing in an individual capacity and/or that their opinions did not represent those of a given organization were also excluded from the analysis. Eventually, this process produced a total of 1,656 testimonies—999 testimonies given by NGOs, 435 by firms, and 222 by expert groups—which constituted the final sample subjected to analysis.1 I take individual witnesses’ organizational affiliations as the basic unit of analysis because the testimonies given mostly address the position of the organization each individual witness represents. I report detailed information about decisions on data conversion scope, group type classification, and text pre-processing in the associated codebook.
In summary, NGOs’ representation at congressional hearings was four times more than that of experts and twice more than that of firms (Figure 1, left). The NGOs’ dominant presence at foreign aid hearings was consistent throughout the period of 1980–2020. However, experts’ representation started to exceed firms’ representation at aid hearings from the early 2000s although with year-to-year fluctuations (Figure 2). It is also noticeable that representation of each of the three groups at aid hearings largely decreased in the mid-1990s. This pattern whereby the number of hearings and the number of testimonies focused on aid peaked in the 1980s and subsequent years and then showed a rapid and significant decline in the mid-1990s (Figure 2) is consistent with Ban et al.’s (2023) recent findings based on all US Congress hearings from 1960 to 2018.2 The authors suggest that the decrease could be attributable to a 1995 reform that led to a massive cut in the number of subcommittees in the US Congress (Ban et al. 2023, 126).
Testimonies by group type at congressional hearings for foreign aid, 1980–2020.
Domestic constituent groups testifying at congressional hearings for foreign aid, 1980–2020.
The NGO and expert groups modestly increased their appearances in the legislative process from the late 1990s onwards, and the firm group maintained its minimal representation at foreign aid hearings with just a few firms presenting testimony in the same period (Figure 2). One possible contributor to maintaining a small number of testimonies given by firms at aid hearings during this period is the OECD's adoption of the Helsinki Package of disciplines in 1991, which led donors to restrict the use of tied aid as a way to promote exports in the global market (Hall 2011). In practice, although the US government had championed the initiative of untying aid globally from as early as the 1970s through the OECD, it was nevertheless the case that US aid had relied heavily on tied aid until the end of the 1980s—mainly because Congress had created tied aid facilities to protect American exporters, serving business interests in the face of strong constituent demands to promote exports.3 It could be that this significant decrease after the late 1990s indicates that Congress had started to fall in line with the international norm for curtailing tied aid practice in their selection of witnesses for committee hearings. It is likely that as private sector witnesses could reasonably be expected to explicitly connect aid to export opportunities, fewer were invited to speak in this changed context. Similarly, firm witnesses may also have become less motivated to gain access to aid hearings for the purpose of lobbying legislators.
Further, differences are evident in the distribution of testimonies among groups depending on whether there was a unified or divided government, as the distribution of witness invitations by group type shows (second plot in Figure 1). A higher percentage of NGOs testified before Congress on aid when there was a Republican majority in the House, and a higher percentage of firms appeared when there was a Democratic majority regardless of whether the majority party had a unified government. The expert group had its highest percentage of appearances when the Republican majority had a unified government.4 Yet, the smallest percentage of experts appeared when a Republican majority with a divided government was in place. Questions, therefore, arise as to why and through what mechanism legislators with different ideological backgrounds select more witnesses from one group type than from others, which is a useful consideration to inform future explanatory work undertaken to explore ideological connections between external groups and legislators.
I analyzed the data to determine how often each group testified before each of the thirty-one House, Senate, and Joint committees comprising the sample (Table 1). Predictably, each group appeared most often in front of the Appropriations Committee. Also, whereas the NGO and expert groups’ witnesses mostly spoke to committees focused on foreign affairs, foreign relations, international relations, and hunger, firm witnesses also spoke on many occasions to the Agricultural Committee and the Banking Finance and Urban Affairs Committee. Supplementary Tables A1 and A2 in the Appendix present descriptive statistics from the data including the entire list of committees in the sample. Further, Supplementary Table A3 in the Appendix describes the five witnesses in each group type with the highest number of appearances before Congress.
Five most common committees for each group type
| Expert | Firm | NGO | ||||
|---|---|---|---|---|---|---|
| Name | Count | Name | Count | Name | Count | |
| 1 | Appropriations | 74 | Appropriations | 148 | Appropriations | 503 |
| 2 | Foreign affairs | 48 | Foreign affairs | 68 | Foreign affairs | 185 |
| 3 | Foreign relations | 38 | Agriculture | 47 | Foreign relations | 111 |
| 4 | International relations | 18 | Foreign relations | 35 | International relations | 50 |
| 5 | Hunger select | 10 | Banking finance and urban affairs | 25 | Hunger select | 45 |
| Expert | Firm | NGO | ||||
|---|---|---|---|---|---|---|
| Name | Count | Name | Count | Name | Count | |
| 1 | Appropriations | 74 | Appropriations | 148 | Appropriations | 503 |
| 2 | Foreign affairs | 48 | Foreign affairs | 68 | Foreign affairs | 185 |
| 3 | Foreign relations | 38 | Agriculture | 47 | Foreign relations | 111 |
| 4 | International relations | 18 | Foreign relations | 35 | International relations | 50 |
| 5 | Hunger select | 10 | Banking finance and urban affairs | 25 | Hunger select | 45 |
Five most common committees for each group type
| Expert | Firm | NGO | ||||
|---|---|---|---|---|---|---|
| Name | Count | Name | Count | Name | Count | |
| 1 | Appropriations | 74 | Appropriations | 148 | Appropriations | 503 |
| 2 | Foreign affairs | 48 | Foreign affairs | 68 | Foreign affairs | 185 |
| 3 | Foreign relations | 38 | Agriculture | 47 | Foreign relations | 111 |
| 4 | International relations | 18 | Foreign relations | 35 | International relations | 50 |
| 5 | Hunger select | 10 | Banking finance and urban affairs | 25 | Hunger select | 45 |
| Expert | Firm | NGO | ||||
|---|---|---|---|---|---|---|
| Name | Count | Name | Count | Name | Count | |
| 1 | Appropriations | 74 | Appropriations | 148 | Appropriations | 503 |
| 2 | Foreign affairs | 48 | Foreign affairs | 68 | Foreign affairs | 185 |
| 3 | Foreign relations | 38 | Agriculture | 47 | Foreign relations | 111 |
| 4 | International relations | 18 | Foreign relations | 35 | International relations | 50 |
| 5 | Hunger select | 10 | Banking finance and urban affairs | 25 | Hunger select | 45 |
Measuring the Evolution of Domestic Aid Interests with Topic Models
The purpose of this study is to characterize the aid topics that three important kinds of domestic groups, i.e., development NGOs, private enterprises, and research experts, have emphasized in their testimonies and how these witnesses’ discussions about aid have evolved over time in the representation process. To uncover common themes in the witness testimonies, I generated topic representations using a topic model called BERTopic. Although traditional topic modeling techniques such as Latent Dirichlet Allocation (LDA) (Blei et al. 2003) are powerful unsupervised tools, their discrete approach to generating latent topics from documents through bag-of-words representations does not allow for an investigation into continuous semantic relationships among words. BERTopic (Grootendorst 2022) extends this topic-generation process by converting each document into its embedding representation using the pre-trained language model Bidirectional Encoder Representations from Transformers (BERT) (Devlin et al. 2018) to obtain document-level information, which is instrumental in producing more accurate representations of words and sentences (Hosseini and Varzaneh 2022).5 The document-embedding feature, in particular, is essential to the analysis, as the text data considered herein consists of a collection of testimonies given by individual witnesses representing the positions of the organizations with which they are affiliated.
To optimize the clustering process, BERTopic takes a further step to reduce the dimensionality of document embeddings before generating semantically similar clusters of documents. In the main model reported in the body text, I employed a technique called Uniform Manifold Approximation and Projection for dimensionality reduction and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) for document clustering. A valuable feature of HDBSCAN is that it does not force all documents to be part of a topic cluster: The documents located far away from the dense regions are treated as outliers and thereby excluded from topic assignments, generated separately as “Topic-1” in the estimation. Lastly, I used class-based term frequency-inverse document frequency (c-TF-IDF) to extract topics from the clusters of documents to estimate topic representations, where BERTopic is primarily distinguished from Top2Vec (Egger and Yu 2022). Both BERTopic and Top2Vec use an embedding approach to document clustering and thus are basically similar in terms of algorithmic structure. However, the c-TF-IDF algorithm in BERTopic allows us to compute the importance that each term has in a topic by generating a set of the most representative terms for each topic (Sánchez-Franco and Rey-Moreno 2022).
I fit the model that generates 15 topics with a set of the 10 most common keywords for each topic using all 1,656 testimonies on aid in the sample. Figure 3 presents the distribution of topics in the documents in two dimensions. As shown, the biggest cluster comprises testimonial documents focusing on food and agricultural aid, followed by the cluster of documents primarily addressing the role of US aid in foreign nations and then the cluster concerning child health. In Figure 3, we can see spatial proximity between multiple clusters of documents, which suggests that the clusters in close proximity have similar semantic characteristics.
Topics in documents: distribution in two dimensions.
These clusters can be roughly grouped into four higher level categories: First, there is a collection of clusters around the center of this plot, which are Food and Agricultural, Role of US Aid, Microenterprise and Credit Union, and Export. Another set of clusters, Environment and Energy are located closely together in the top-left area. Multiple clusters broadly related to the socio-economic aspects of aid such as Child Health, Nutrition, and Education are grouped together in the bottom-left area. Lastly, cross-cutting issues related to aid, including Peacekeeping (South Sudan, Africa), Humanitarian and Refugee, Democracy and Election (Egypt, MENA), and Population and Family Planning are located in close proximity in the center-bottom area. A single cluster on the Military in Pakistan–Afghanistan relations is located in the right-bottom area, slightly distanced from the rest of the clusters. Figure 4 then presents the ratio of individual topic representations in Figure 3, coupled with a set of the ten most common words for each topic. The top three topic clusters take up more than 50 percent of the entire topic distribution of aid testimonies, whereas the topic clusters with a focus on specific countries such as South Sudan, Egypt, Pakistan, and Afghanistan account for a relatively small portion of the overall topic representations.6 In Appendix 3.1, I explore how the discussions of domestic groups differ between Senate and House committees. I present the relative proportions of each topic representation in Senate committees with those in House committees set as 100 percent. The results indicate that topics such as Military in Pakistan–Afghanistan and Democracy and Election (Egypt, MENA), which are of strategic importance and related to national security, tend to be favored by Senate committees, while topics such as Education and Nutrition are generally characterized by higher engagement of House committees.
Topics in words: ratio and most common words for each dimension.
Next, I trained the model to extract topic representations for each group type in order to explore whether and how NGO, firm, and research expert witnesses approach common topics differently. First, as shown in Figure 5, the five topics most frequently referenced by firm witnesses comprise (1) Food and Agricultural, (2) Export, (3) Role of US Aid, (4) Energy, and (5) Microenterprise and Credit Union. As one might expect, the firm witnesses focused on the set of topics directly related to the economic aid sector, such that the remaining fifteen topics were represented to a far lesser extent by firms than by either the NGOs or the research experts. For example, no speakers from the private enterprise side testified in relation to either Education or Nutrition.7
Topic representation for each group type: expert, NGO, and firm.
The four most frequent topic representations for NGOs and experts were the same: (1) Role of US Aid, (2) Food and Agricultural, (3) Child Health, and (4) Environment. The fifth most frequently referenced topic for NGO witnesses was Humanitarian and Refugee. For expert witnesses, Military in Pakistan–Afghan Relations and Microenterprise and Credit Union tied for fifth. There are some differences in topic representations between the NGO and expert groups: For example, Military in Pakistan–Afghan Relations was the topic referenced least often in the NGO group's testimonies but frequently referenced by the research expert group. The Education, Energy, Humanitarian and Refugee, and Nutrition topics were seldom mentioned by the research experts but frequently referenced by the NGOs.
Supplementary Table A8 in the Appendix presents topic representations generated for each of the five most frequently-appearing witness groups from the NGOs, firms, and experts, coupled with excerpts from each witness's testimony. In Appendix 3.3, I report and explain in greater detail the key agenda each group presents in the testimonies given. The result shows that the ways individual witnesses appeal to committees clearly differ according to group type. In summary, it is particularly interesting to note that the frequently-appearing NGOs and firms not only advocate for changes in policy and legislation as a way to support their own activities in relation to aid but also make budget requests—e.g., for larger subsidies for their projects or increases in the AID budget in their aid sector—to benefit their operations both directly and indirectly. On the other hand, most of the frequently-appearing research experts focused primarily on the overall direction of the US aid program, with their views varying depending on the kinds of goals and missions each witness pursues, e.g., aid as a long-term investment in promoting the political security and economic interests of the United States or aid aimed at helping the poor and promoting growth in recipient countries.
More detailed discussions of the substantive findings on witnesses’ advocacy narratives and approaches are provided in Appendix 3.4. First, I examine the differences in witness representations between Democratic and Republican majorities in the House in Figure 1. Here, I present six different topic frequencies for each cluster, distinguished by group type and whether they occur under Democratic or Republican control. I also classify the legislative discourse of NGO, firm, and expert witnesses into eight different categories to describe their advocacy narratives and approaches.
I then fit the model to compute the proportion of topic representations for 10-year periods, which allows us to look at the evolution of aid topics in US politics over time (Figure 6). The focal period of the study includes a series of exogenous events and phenomena that almost certainly had a significant influence on discussions about foreign aid in the United States. These include the Cold War, globalization, the adoption of the United Nations Millennium Development Goals (MDGs) and Sustainable Development Goals (SDGs), and wars in Iraq and Afghanistan. It is, thus, reasonable to expect witnesses’ testimonies before the end of the Cold War to differ from witness testimonies after the end of the Cold War in terms of the central aid topics covered. Even when witnesses testifying in different time periods address the same topic, the emphasis or the approach may reflect differences depending on the time period.
Topics representation per time period: 1980–1990, 1991–2000, 2001–2010, and 2011–2020.
During the first period, i.e., 1980–1990, “Population and Family Planning” is the most prevalent topic with witnesses arguing that the United States should provide more assistance to meet the demands of million couples in the developing world in need of family planning (Figure 6). The second most prevalent topic during this period is “Export,” with the majority of testimonies on this topic given by firm witnesses trying to persuade Congress and the government that the United States should provide more support for domestic exporters using aid.
In the 1990s, Energy became the most prevalent topic for aid. Many witnesses advocated for the importance of improving energy efficiency and renewable programs through US aid efforts. The Humanitarian and Refugee topic emerged in the 1990s and immediately become salient in the testimonies of all three groups. Also, discussions about Aid to Africa are more frequent during this time period than in any other period examined, which witnesses tended to address as part of the aftermath of the end of the Cold War.
Substantial changes took place in terms of overall topical prevalence in the period of 2000–2010. Several topics highly salient in the 1980s and 1990s such as Export, Environment, and Energy were addressed only very minimally during this time, and discussions on Population and Family Planning even virtually disappeared. In contrast, new topics relating to socio-economic development such as Education and Nutrition became highly prevalent, in addition to greater emphasis on Democracy and Election issues.
In the period of 2010–2020, Peacekeeping (South Sudan, Africa) became the most prevalent topic, followed by Education, Democracy and Election (Egypt, MENA), and Military in Pakistan– Afghan Relations, indicating that discussions about aid during this time period became more diverse in relation to the contexts of individual countries. It is also worth noting that topics such as and Role of US Aid and Food and Agricultural were generally salient throughout the whole time period.
In Supplementary Figure A9, I display the evolution of aid topics estimated through Dynamic Topic Modeling. While Figure 6 presents the relative prevalence of topics across 10-year periods, Supplementary Figure A9 depicts changes in each topic frequency over time. Particularly, it shows increases in topic representations directly related to socio-economic development advocated by the MDGs and SDGs,8 particularly after 2000. For example, the Child Health topic indicates a rapid increase in early 2010, peaking in 2013—a notable rise, especially considering the overall decrease in the number of testimonies since the mid-1990s. The topics of Nutrition and Education began to emerge in witnesses’ testimonies in the late 1980s and the 1990s, respectively, with the proportion of Education increasing in the late 1990s, reaching its peak in the late 2000s. Overall, the result suggests how global norms and standards could potentially be internalized in legislative discourse.
Discussion and Conclusion
In this paper, I examined US congressional testimonies as one way to characterize interactions between legislators and domestic interest groups in US aid decision-making. Based on new data as presented, my findings pertain to how the composition of witness groups testifying at aid hearings and the aid topics discussed by those groups in that context have varied over time. My goal was to offer a descriptive overview of domestic aid discussions in the United States over the last four decades to build a foundation for future research. However, my findings go further by providing some empirical support for the critical role of domestic politics in the aid decision-making process.
First, my results provide quantitative evidence consistent with the idea of the distributive politics of aid (Roberts 2014; Milner and Tingley 2015). Individual witnesses invited to hearings advocate for or argue against specific legislation, policies, and budgets in order to advance their own agenda in relation to aid. The topic representations generated for each group type and for each time period show that NGO, firm, and expert witnesses differ both in terms of the sectoral/topical interests they represent but also in terms of the advocacy strategies they favor in the aid decision-making process. More specifically, I found that NGOs and firms testifying at aid hearings not only normatively advocated for or against changes in the state's aid policies, laws, and budgets but also lobbied for financial support for their own aid work. Here, whereas the majority of private companies focused primarily on the use of US aid to support US exporters in the global market, the majority of NGOs emphasized the various impacts of US aid on the socioeconomic development of recipient countries. In comparison, research experts focused less on making budget requests and more on providing analytical information on the strategic and effective use of US aid. However, their views as to whether US aid should focus on promoting the country's own political security and economic interests or on reducing poverty and promoting growth in developing countries, or take both concerns into account tended to diverge depending on each expert group's goals and activities.
In addition, my results show that the appearance of private companies at aid hearings declined substantially in the late 1990s. A decrease in the number of testimonies given was evident for each witness group type during this period, arguably due to the reform of subcommittees in 1995. However, a modest recovery was found for the NGO and expert witness groups after this period whereas only a few firms testified from the late 1990s onwards. As previously stated, it is likely that the large decrease in firms’ testimonies is related to the OECD's adoption of the Helsinki Package of disciplines in 1991, which led to restrictions on the use of tied aid in promoting exports in the global market. This account is also consistent with my finding that the topic representation for Export all but disappeared from testimonies from 2001 onwards. While such decrease in firms’ appearances at aid hearings is striking, does this mean that private companies entirely gave up on appealing to legislators on promoting their exports and trade in the global market in connection with aid? Future research could explore this question further through an analysis of the language companies used to lobby legislators for foreign aid and trade/exports based on Lobbying Disclosure Act data and compare the results before and after the late 1990s.
Lastly, my descriptive finding shows that a higher percentage of NGO witnesses testified when there was a Republican majority in the House rather than a Democratic majority whereas the opposite held for firm witnesses. The expert group had its highest appearance rates under the Republican majority with a unified government. Although foreign aid is considered a bipartisan issue and not particularly salient in domestic politics, there may be some ideological connections that influence legislators’ decisions on which witness groups to invite. It would be interesting for future research to explore the reasons for and the conditions in which liberal and/or conservative ties drive legislators to select more witnesses from one group type than from others.
Overall, my analysis reflects the evolving contexts of US aid over time, providing an empirical foundation for future research on how changing discourses and external events shape US policymaking. The evolution of aid topics correlates with significant global shifts such as the end of the Cold War, after which their prominence in legislative discussions largely declined. The 2000s marked a shift in focus to topics such as military assistance, peacekeeping, and democracy, as well as an increased emphasis on regions such as Africa and the Middle East. Beyond discussions of aid as a diplomatic tool, topics related to socio-economic development also gained prominence during this period, indicative of the advocacy efforts by NGOs and research experts aimed at steering US aid toward promoting global poverty reduction and sustainable development.
Author Biography
Esol Cho is an Assistant Professor of International Studies at Xi'an Jiaotong-Liverpool University. Her research focuses on international political economy and the domestic politics of foreign aid.
Footnotes
I present the results from the witnesses’ spoken testimonies in the main text. The sensitivity analysis is conducted against different samples that include only written testimonies or both oral and written testimonies, as provided in the Appendix.
Figure 1 in Ban et al. (2023), which visualizes the frequency of US congressional hearings over time, indicates the same distribution.
For example, under the 1983 Trade and Development Enhancement Act, Congress granted Ex-Im the authority to provide loans at interest rates lower than its own borrowing costs, enabling it to align with foreign credits (Hall 2011, 662).
The US think-tank literature notes that their proliferation since the 1980s can be attributed to an increased demand for their role as policy evaluators, a response to partisan polarization in Congress. This work suggests that their prevalence may be partially explained by conservative ties, which amplify conservative positions in US policymaking through shared political visions and personal connections. See, e.g., Bertelli and Wenger (2009) and Garsten (2021).
In the Appendix, I provide a detailed explanation of the benefits and limitations of this method compared to other computerized text analysis software.
I provide the results of robustness checks for the topic representations in Appendices 2.1 and 2.2. Specifically, I investigate how various issues, such as “aid and the Cold War” and “democracy aid,” are distributed across different topic clusters, which are estimated to be less prevalent than originally expected.
A question may arise considering the significant presence of firms in the agribusiness sector that could suggest a connection with the topic of Nutrition. I have checked for the robustness of the results, as detailed in Appendix 2.3.
The international community launched the MDG campaign in order to collectively push foreign aid toward fulfilling the basic needs of recipient countries with an emphasis on eradicating extreme hunger, halting the spread of HIV/AIDS, improving maternal health, reducing child mortality, and advancing toward universal primary education. The United Nations adopted the MDGs in 2000 and later the SDGs after the target year of 2015.
