A Push Toward Autocracy? The Effects of US Security Cooperation on Democratic Development in Africa

Introduction

In the post-Cold War era, the overarching foreign policy of the USA has comprised two key pillars: first, the promotion of democracy across the globe; and secondly, the provision of military assistance to allies and friends to support the USA in achieving its strategic defense priorities. Although US governments often spoke about the importance of democracy and development in the aftermath of World War II, they did not heavily promote, or fund, democracy promotion programs until the end of the Cold War. Beginning with the Clinton administration in the 1990s and continuing throughout the Biden administration, the USA treated global democracy promotion as a major part of its approach to international relations (Dalacoura 2005). From 2001 to 2024, the USA spent between US$1.18 billion and US$4.16 billion, annually, on democracy, human rights, and governance programming around the world (ForeignAssistance.gov).

The US approach to military assistance, which focuses on building the capability and capacity of partners, has also spanned multiple administrations under different parties (Omelicheva, Carter, and Campbell 2017). The USA had used military assistance to support coalition partners during the 1990s as part of the Gulf War, but after the terrorist attacks of September 11, 2001 it dramatically increased its military assistance to a variety of countries to build partner capacity and help fight terrorism (Sullivan, Tessman, and Li 2011). That year, the USA provided US$4.35 billion of military assistance to its global partners, which represented about 30% of its total foreign assistance (ForeignAssistance.gov). In absolute terms, this figure increased significantly over the next decade. In 2011, the USA spent US$16.84 billion on military assistance, about 35% of its annual foreign assistance that year (ibid).

Given that the USA has simultaneously pursued both democracy promotion and the improvement of its partners’ military capacity, often in the same country, how have these two tenets of foreign policy imperatives interacted? Are they mutually supportive or have they worked at cross-purposes? This paper examines the relationship between Africa’s democratic progress and US security assistance to the region from the mid-2000s to 2020. This assistance encompasses a variety of defense partnership activities including military assistance. Through a quantitative analysis of trends over time, we are able to link security assistance, democracy assistance, and democratic outcomes.

Why do we examine Africa in particular? These two trends – rapid democratization and an exponential increase in security assistance – are especially pronounced in that region. The third wave of democracy affected much of Africa in a relatively short period of time. During the 1990s, roughly 75% of African countries began to transition to some form of democratic rule. Then, with the proliferation of terrorist threats in the 2010s, security assistance increased markedly. In 2000, the USA provided approximately US$50 million in military assistance to the African continent.¹ In 2010, this figure increased to about US$350 million, and by 2020 this amount had increased to at least US$544 million (Foreign Assistance.gov). Thus, there are a number of relatively new democracies in Africa, many with shallow roots, that have received a substantial influx of military assistance over the past two decades. Additionally, the timeframe – 2005 to 2020 – encompasses several relevant factors: the rise of violent extremist movements (2006); the creation of the US Africa Command (AFRICOM) (2007); and new and persistent challenges to democratic governance, including electoral violence, the rise in social and protest movements, the decline in political rights and civil liberties, and the gap between supply and demand of democracy in Africa.

Our analysis of democracy promotion and security assistance, from both a theoretical and an empirical perspective, reveals that military assistance, in particular, can actively undermine democracy promotion, unless specific care is taken to counterbalance these effects. Although this may not be the intention of US military assistance, it is a risk that the USA takes when pursuing multiple, competing goals in a given country. As such, the findings of this paper have significant implications for US foreign assistance and the ways in which external partners may impact a country’s democratic trajectory.

Democracy Promotion and Military Assistance: Mutually Supportive or Inherently at Odds?

US commitment to global democracy promotion has a long history. However, the impact of US foreign assistance on democratic outcomes has been a source of debate, contradiction, and nuance. Direct assistance for democracy focuses on support to key actors and institutions necessary for democracy such as political parties, the judiciary, election management bodies, civil society, and the press. The rationale has been that strengthening the different facets of democracy will, in turn, bolster and deepen the overall quality as well. According to Gisselquist et al. (2021), targeted democracy assistance has had a positive impact on democratic development. However, Finkel et al. (2020) find that targeted democracy assistance is most successful in countries with lower levels of military assistance. Similarly, Savage (2017) argues that democracy assistance fails in countries with larger-sized militaries. He attributes this failure to the relatively “privileged position” the military tends to hold in authoritarian regimes and their ability, due to their large size, to thwart attempts at political liberalization.

Scholars and practitioners also argue that economic development assistance should have a positive, albeit indirect, impact on democracy as many have found a correlation between economic development and democratic development (Lipset 1959; Heo and Tan 2001). The exact nature of this relationship (unidirectional, interdependent, or spurious) is also a frequent source of debate in the literature (Bueno de Mesquita and Downs 2005; Robinson 2006; Gisselquist et al. 2021). Looking at Africa specifically, with data largely from the Cold War period, Goldsmith (2001) found that economic assistance had a positive effect on democracy, although there were very few democracies in Africa during most of the time period he analyzed. Kalyvitis and Vlachaki (2012), on the other hand, when looking globally, found that higher levels of economic support have had a negative impact on democratic development, but that when countries undertook steps towards economic liberalization, this relationship disappeared. Examining data from 1988–2001, Scott and Steele (2011) analyze democratic programming alongside economic assistance and find that specific democracy promotion packages did have a positive impact on democratic quality whereas economic aid had no effect.

Turning to the impact of military assistance, studies conducted during the Cold War generally found no impact of military aid on a country’s democracy, as measured by protection of human rights (Cingranelli and Pasquarello 1985). According to Travis (1995), human rights played little to no role in military assistance decisions in the 1980s. But he also argued that the USA was sending mixed messages about its commitment to democracy promotion: on one hand, the USA withheld economic support assistance to punish countries accused of human rights abuses, yet, on the other hand, military assistance was awarded regardless of a country’s human rights record. It seems that decisions about the provision of military assistance were largely either oblivious or agnostic to the potential impact on democratic development. Military assistance decisions are generally based on a combination of partner needs and willingness, and US strategic considerations. Larry Diamond (2021) argues that the strategic decisions and partnerships the United States made as part of the Global War on Terror (GWOT) have been a major contributing factor to global democratic backsliding. Diamond argues that the United States prioritized support in the GWOT over most other things, which in turn has enabled autocratic rulers to maintain and deepen control. Although an indirect effect on democratic development, Savage and Caverly (2017) found that US military training, in particular, increases the likelihood of a coup attempt taking place in a given country. They argue that the increased professionalism that military training provides to partner nations results in improved “human capital” that can then increase the power and capabilities of the military relative to the government or regime in power. As a successful coup generally leads to an interruption in constitutional government, democratic quality tends to suffer as a result.

US Security Partnerships in Africa

The United States is concerned about a constellation of security threats in Africa that includes piracy in the Gulf of Guinea and off the coast of Somalia and terrorist activity in the Sahel, Lake Chad Basin region, and the Horn of Africa. Over the past two decades, the United States has dramatically increased its security cooperation to African partners. Since 2001, the United States has provided at least US$5 billion in security assistance to a variety of African partners to help address a variety of security threats.² Between 2009 and 2015, the Department of Defense (DoD) increased spending to the AFRICOM Area of Responsibility (AOR) by an estimated 650%, notifying Congress of more than US$400 million in Title 10 spending – security cooperation programs – in fiscal year (FY) 2016, near the peak of military assistance to Africa (Husted et al. 2020). For FY21, the United States budgeted an estimated US$227 million for security cooperation programs and activities in AFRICOM’s AOR (United States Government, Office of the Secretary of Defense 2020).

Military assistance generally refers to training programs and equipment provided to a partner’s military forces as authorized by US statutes. However, security cooperation entails much more than just military assistance. The DoD defines security cooperation as:

Security cooperation includes joint exercises, military basing and overflight agreements, and assist and accompany missions, as well as programs to build partner capacity and strengthen defense institutions, which include the purchasing or transfer of defense equipment.

The US government engages with African countries using various types of security cooperation activities for multiple reasons, not all of which are purely operational. Some assistance is motivated by the need for access and influence while some is directly related to improved security outcomes. Political motivations include: access for basing or operating; support for regional operations and peacekeeping; projecting power against global rivals; and support in international disputes. Defense motivations include building partner capacity and capability to counter threats that may eventually reach the United States. More specifically, the intended operational outputs of US military assistance to partner nations include increased security force capabilities and capacity, increased respect for human rights by security forces, strengthened defense institutions, and increased military professionalism. In the long term, strategic outcomes of military assistance include decreased trafficking and increased interdictions; decreased terrorist group strength, support, and attacks; and increased perception of individual citizen security in partner nations.

To fully understand the impact of security cooperation, it is important to look at a variety of factors in addition to the monetary value of security assistance provided to a partner. Additionally, due to the imprecise nature of security assistance data, especially DoD-reported data, it is preferable to include data from diverse sources for a variety of indicators to better estimate the full impact of US security cooperation (Grill et al. 2017). We collected data on different aspects of the security relationship between the United States and partner countries as a way of triangulating impact and best representing the intensity of US interest in a given country. We collected data on three different annual indicators that reflect the importance of the USA’s security relationship with a partner country: the number of DoD personnel deployed in a country; the total number of foreign troops trained; and the total amount of military assistance disbursed. Each indicator and its incidence in Africa are discussed in depth below. Most data is available beginning in the mid-2000s, so we analyzed data from 2009 through 2020, or when the most recent data are available, for a total of 53 African countries.³ These years cover two different administrations (Barack H. Obama and Donald J. Trump), one Democrat and one Republican, and five different combatant commanders.

DoD Deployments

The DoD reports the number of military and civilian personnel permanently⁴ assigned duty by US state and country on a monthly basis and makes this data publicly available. The US government has a small military footprint in Africa compared to other regions. During the timeframe that we analyze, this consisted of: one permanent base, Camp Lemonier in Djibouti; two Forward Operating Sites; 13 Cooperative Security Locations; and 12 Contingency Locations in countries including Senegal, Ghana, Gabon, Uganda, and, Kenya (AFRICOM 2019). On average, there were about 2,100 active-duty military service members deployed to Africa in 2010 (DoD Defense Manpower Data Center). This figure increased steadily to more than 6,400 troops deployed in 2017 and then precipitously declined to less than 2,000 in 2018, and less than 1,000 troops in 2019 and 2020. With between 22,000 and 48,000 overseas deployments annually, the number of troops stationed in Africa at any given time over the period 2009–2020 has never been greater than 2.3% of all its international deployments (Ibid). Countries with the largest number of US troops include Niger (554 in 2017); Somalia (289 in 2017); Ethiopia (143 troops in 2011); Kenya (141 troops in 2012); Tunisia (106 in 2017); and Libya (105 in 2013). Table 1 and Figure 1 provide continental trends from 2009 to 2020.

Table 1

DoD Troops Deployed in Africa.

YEAR	NUMBER OF TROOPS IN AFRICOM’s AOR	PERCENT OF GLOBAL DEPLOYMENTS IN AFRICA	TOP 3 COUNTRIES (NUMBER OF TROOPS IN PARENTHESES)
2009	2215	0.46	Djibouti (1798); South Africa (108); Kenya (73)
2010	2171	0.48	Djibouti (1791); South Africa (134)
2011	3007	0.67	Djibouti (2215); Somalia (182); Ethiopia (143)
2012	3705	1.00	Djibouti (2764); Somalia (174); Kenya (141)
2013	3878	1.15	Djibouti (2623); Somalia (314); Kenya (135)
2014	3907	1.27	Djibouti (2827); Niger (120); Kenya (118)
2015	4928	1.74	Djibouti (3643); Somalia (237); Niger (157)
2016	4052	1.48	Djibouti (3406); Somalia (107); Uganda (84)
2017	6463	2.22	Djibouti (4715); Niger (554); Somalia (289)
2018	1931	0.85	Djibouti (1346); Somalia (60); Kenya (49)
2019	665	0.29	Djibouti (91); Somalia (85); South Africa (49)
2020	846	0.38	Djibouti (179); Somalia (73); Kenya (55)

Figure 1

US troops are deployed to foreign countries for a number of reasons: to provide training or operational support in response to an immediate threat, to signal US alignment to potential adversaries as a means of deterrence, or to project power to both allies and adversaries. Recent research has found that US troop deployments are associated with regime stability in the host country (Braithwaite and Kucik 2018) and a decreased likelihood of coup attempts (Allen et al. 2023). The presence of US troops tends to convey economic benefits to the host country and can act as a strong indication that the USA supports the incumbent regime. Research on the impact of US deployments on democratic development does not exist. However, by increasing regime longevity, US troops may be inadvertently propping up non-democratic regimes. Based on existing literature, we would expect, in the long term, an indirect but potentially negative impact of US troop deployment on democratic development.

Could it be that US troops are deployed only to countries with security threats and these countries tend to be less democratic? Looking at troop deployments trends in 2008, 19 African countries had no US troops deployed within their borders, while 23 African countries did. Of those 19 countries with no US troops, only one of them had an active conflict.⁵ In the 23 countries with US troop deployments, almost half (44%) had an active conflict that year. Between the two categories of countries, those with US troops and those without, there was no significant difference in the quality of their democracy. The average Electoral Democracy Index score was 0.42 for those without US troops and 0.42 for those countries with a US troop presence in 2008.

But this relationship changed dramatically by 2018, both in terms of the number of countries that had a US troop presence and which countries had active conflict going on or not. In 2018, 11 countries (or less than 20%) hosted no US troops that year, whilst the majority of countries did. In addition, of the 11 countries without a US troop presence, 6 of them (55%) had an active conflict. Of the countries that did have a US troop presence that year (42 or around 80%), only 19% were experiencing an active conflict. Thus, while it may have been initially true that the USA was deploying troops only to countries with an acute security need, by the late 2010s, this was no longer the case. The USA was deploying broadly across the continent, albeit in small numbers, and mostly to countries without an active conflict. Comparing the two groups of countries, in 2018, those without US troops had a slightly higher average democracy score than those who hosted US troops, 0.48 versus 0.43, respectively.

This leads to our first hypotheses:

Hypothesis 1a: The number of US troops deployed in a country is negatively correlated with democratic development in the host country, all things being equal.

Hypothesis 1b: As the number of US troops deployed in a country increases, government accountability decreases.

Military Assistance

This measure captures training and equipment provided by the United States to partner countries. Both the State Department and USAID maintained online databases of assistance. These two sources presented data in different ways and cover different time periods. The USAID version broke down assistance into two categories: economic and military. Figure 2 details the amount of military assistance reported as disbursed to African countries annually from 2009 to 2020. This data reportedly includes both State Department- and DoD-funded programs, but there are many suspected gaps.

Figure 2

The United States provides hundreds of millions in military assistance to African countries and this amount has been increasing since the late 2000s. In 2009, the USA provided roughly US$265 million in military assistance to African countries; in 2019, this number had more than doubled to US$627 million. Again, the amount of assistance provided both from year to year and across countries varies widely. Some countries, such as Equatorial Guinea, receive little to no military assistance while others receive significant assistance. Kenya, Uganda, Chad, Niger, and Tunisia, for example, have very large programs that amount to tens of millions in military assistance per year.

The large monetary values associated with military assistance reflect the costs of military equipment, weapons, weapons systems, and the logistics of providing these items, often accompanied by maintenance packages. Providing such large influxes of high value materiel could theoretically have a deleterious impact on the quality of a country’s democracy. High-value security assistance frees up government spending that can be used for clientelistic or other corrupt purposes and can enable governments to maintain power by securing the support of a select segment of the population, namely elites. Alternatively, security assistance, as an external source of revenue and materiel support, can render governments unaccountable to popular demands and pressures.

We hypothesize that the amount of military assistance the USA provides to partner countries can have a negative impact on democratic development, depending on the totality of the US engagement. If the US engagement is primarily militarily or security focused, we believe the relation between military assistance and democratic development should be negative. However, if the US engagement includes increased amounts of economic support and/or democracy support, we believe this relationship should be positive. Thus, our final hypotheses are:

Hypothesis 3a: As the proportion of US military assistance increases relative to all other forms of foreign assistance, the quality of a partner country’s democracy decreases, all things being equal.

Hypothesis 3b: As the proportion of US military assistance increases relative to all other forms of foreign assistance, government accountability decreases.

Results

When looking across the different aspects of security cooperation, some trends emerge. We find, not surprisingly, that many of these measures are only very loosely correlated. Table 2 provides correlation statistics for key independent variables.

The only negative relationship is that between the number of African military troops trained per year and the number of DoD personnel deployed in Africa, but again, the relationship is very weak. DoD deployments are generally not strongly related to any of the other aspects of security cooperation. This may be because the number of US personnel deployed to Africa is so small that it does not represent any genuine priorities. This may also reflect AFRICOM’s “by, with, and through” approach to African conflicts, in which the United States prefers to provide support to partner militaries as opposed to directly intervening (Vergun 2020). The very weak relationship between these three aspects of security cooperation could also reflect the multiplicity of actors responsible for making decisions about who the United States should partner with in security matters and how. For example, the Office of the Secretary of Defense, AFRICOM, the State Department, and at times the National Security Council all have different roles to play in partnership decisions.

Our multivariate regression analysis examines the relationship between US security cooperation and democracy and governance. As each of the three security cooperation variables is attempting to capture a different aspect of the relationship between the United States and a partner nation, we analyzed each relationship separately using the appropriate control variables and model structure. We are using cross-national time-series data, where years are nested within countries, which requires that we take into account the presence of time-invariant variables and the dependent nature of the residuals produced within each country. We use Random Effects models to address these issues, where the standard errors are clustered by country.

We include the current time period time lags of one, two, and three years for each of our independent variables because we believe that the effects of security assistance on democracy and governance are unlikely to be immediate. To effectively train partner security forces with a sustained capability, for example, takes multiple years of engagement, training of several units, and ideally, institutional capacity support. Additionally, US security cooperation is generally reported by FY, which creates a temporal disconnect between the provision of assistance (which runs from October to September) and annual indices (which typically align with the calendar year). Finally, although the time lags suggest a causal relationship, we are mindful that these are not sufficient to demonstrate causation and that we are reporting on associations that could potentially be explained by other variables.

We used multiple measures of democracy and governance both as a robustness check to ensure that we fully captured what democracy entails and to help identify which aspect(s) of democracy the security cooperation could be affecting. Specifically, we measured democracy and governance with data on the freeness and fairness of elections; respect for political rights and civil liberties; and the quality of governance.

To measure the extent that elections are free and fair and produce accountable governments, we used data from the Electoral Democracy Index. The index ranges from 0 to 1; 0 represents a country where electoral democracy is not achieved and 1 represents a county where democracy is fully achieved (Coppedge et al. 2021a; Coppedge et al. 2021b). For this indicator, the focus is on the responsiveness of the government to citizens through electoral mechanisms. While freedom is considered to some extent, factors like freedom of media and freedom of expression are linked more to electoral outcomes than to other normative conceptions of liberal democracy.

We used data from Freedom House, a non-profit organization that historically has received much of its funding from the US government. Freedom House’s annual Freedom in the World report provides a numeric score for more than 200 countries and territories in the world. The score is the product of an analysis of 10 political rights indicators and 15 civil liberties indicators as assessed by a team of country experts (Freedom House).⁶ Countries with lower scores (1–2,5) are considered Free; countries with moderate scores (3–5) are considered Partly Free; countries with the highest scores (5,5–7,0) are considered Not Free.

Finally, we included data from the Mo Ibrahim Index of African Governance (IIAG). This provides an annual rating of the overall quality of a country’s governance based on more than 100 factors broken down into four categories: security and rule of law; participation, rights, and inclusion; foundations for economic opportunity; and human development. Of most relevance to this study, we examined sub-categories of data from IIAG on accountability and transparency. Scores range from 0, indicating no governance capacity, to 100, representing high quality of governance.

Using different measures of democracy and governance helps us to get some purchase on the specific mechanisms that could be driving the relationship between security cooperation and democratic development. As with the descriptive analysis, our data ranges from 2009 to 2020 and cover 53 countries in Africa, for a total of more than 500 observations, depending on the specific model. Data availability reduces the total number of cases and countries in some models. Table 3 reports the summary results of all our models tested. Appendix A has more detailed results of our statistical analysis. We discuss each key security cooperation variable in turn below. We note that the relationships show correlation, but do not show a definitive causation.

Discussion

This article investigated the relationship between US security cooperation activities, and the different facets of assistance that this includes, and key indicators of democratization in Africa between 2005 and 2020. The timeframe is shaped by the availability of specific data, but encompasses a number of key developments in Africa’s security landscape that have shaped US–Africa policy: namely, the creation of AFRICOM in 2007 and its formal establishment in Germany in 2008, and the rise of violent extremist movements in the mid-2000s. At the same time, democratic governance faced new challenges and persistent old ones. For example, electoral violence, which had been expected to decline with the adoption of democratic norms, remained stubbornly constant. In 2005, social and protest movements began to rise sharply in Africa and, after a steady rise, civil liberties and political rights began to decline in 2006. Finally, demand for democracy has consistently outstripped its supply by African governments (Mattes 2019). In the context of the timeframe’s fluid security and governance context, we examined the relationship between democratic trends and high-level security attention on countries through the number of DoD personnel deployed in a country, the number of foreign troops trained, and the amount of military assistance disbursed. Democratic indicators measure good governance, through the IIAG; civil liberties and political rights, through Freedom House; and free and fair elections, through the Electoral Democracy Index.

Overall, the relationship between US security cooperation and democratic indicators in Africa has a mixed, albeit generally moderately negative, track record. The number of DoD personnel deployed and the number of foreign troops trained are generally not consistently correlated with democracy and governance indicators. However, the greater the amount of US military assistance given to African countries in relation to other types of assistance, the stronger the negative impact on protection of freedoms and civil liberties and on government accountability. In other words, overmilitarized relationships may present challenges to advancing democracy and achieving a key tenet of US foreign policy in Africa. However, it is important to note that while our findings of the effects of US security assistance demonstrate correlation and suggestion, we do not conclude that they are monocausal relationships.

This research contributes to our understanding of how the pursuit of security cooperation partnerships, and the provision of military assistance in particular, may result in short-term security gains but at a significant long-term cost to democratic development. A myopic focus on military assistance, with insufficient attention to other facets of international assistance that are indirectly related to the need for military assistance, such as economic development or support for democracy, may result in an unbalanced approach to international assistance, or overmilitarized relationships where the security partnership is prioritized over all else. Although a country or region may experience immediate improvements in their security situation through the stabilization of an acute crisis or the enhanced ability of military professionals to address ongoing insecurity, in the absence of deliberate steps to protect democracy, the long-term impacts of military assistance may be negative and, ultimately, counterproductive. From a foreign policy perspective, addressing today’s security crisis too narrowly may result in a contraction of the democratic space, prompting new grievances against a state, which can then contribute to additional instability and insecurity in the future.

Additional File

The additional file for this article can be found as follows:

Competing Interests

The authors have no competing interests to declare.