Sage Journals: Discover world-class research

Abstract

Why do coups happen and how can state leaders affect their likelihood? Existing research focuses on structural factors as well as “coup-proofing” as drivers of coup risk. I argue that the literature misses an important alternative avenue by which leaders affect the likelihood that their militaries remove them from office: adding or removing military officers from the government. When leaders bring military officers into the government, they signal to the military that there is a peaceful path to sharing power and provide an alternative to coups. In contrast, removing military officers from government dramatically increases the risk of a coup attempt as the military retaliates against the leader’s power grab. I test this theory using cross-national data that captures changes in military representation in national cabinets and state councils from 1969 to 2008 and find mixed support. In line with the theory, there is strong evidence that large reductions in the military’s government positions lead to coup attempts, and more modest evidence that large increases reduce coup attempts. However, contrary to expectations, the results also suggest that small increases in military government positions increase the risk of a coup.

Keywords

Coups civil-military relations power-sharing

Why do coups happen and how do state leaders affect their likelihood? Coups remain a major mechanism by which irregular leader changes occur (Geddes, 1999), and 2021 and 2022 have seen a sharp uptick in the number of coup attempts.¹ The military-led regimes that often result from successful coups have a much higher propensity for international conflict (Weeks, 2012) and are more likely to repress political opponents (Poe et al., 1999).

There is a well-developed literature on the causes and consequences of coups as well as “coup-proofing.” Coup-proofing includes building rival security organizations, overlapping command hierarchies, and “ethnic stacking” of the military (Quinlivan, 1999; De Bruin, 2020). Less understood is how the military’s role in government affects coups. This should be a critical factor in the coup propensity of all regimes where the military and political leadership are distinct actors. I argue that bringing military officers into the highest levels of the government limits the leader’s power, shares power with the military, and forestalls a coup. In contrast, the removal of military officers from government introduces a great deal of instability to government-military relations. The military loses its role in policy-making, the leader’s power is increased, and distrust between the government and the military increases.

I test this theory with data on coup attempts and the number of cabinet-level government positions held by military officers and find mixed support. I find strong evidence that large net reductions in the number of military officers in government substantially increase the risk of coups and weaker evidence that large increases reduce coup risk. However, there is evidence that smaller increases and decreases in military participation in government both increase coup risk.

Coups and coup-proofing

Major drivers of coup risk include economic development, (Collier and Hoeffler, 2005), weak civil society (Belkin and Schoffer, 2003), crises in civilian politics (Finer and Edward, 1973), ethnic conflict (Harkness, 2018), officer grievances (Albrecht and Eibl, 2018; White, 2020), and security environment (Piplani and Talmadge, 2016). In response to coup risk, leaders coup proof. This can include multiple command hierarchies, loyalist paramilitaries, or promotion criteria based on political loyalty (De Bruin, 2020). Leaders may also engage in ethnic stacking of their militaries (Quinlivan 1999). Coup-proofing is generally associated with a reduction in coup risk (Powell 2012). However, it comes with significant trade-offs in terms of military efficacy (Talmadge, 2015) and civil war risk (Roessler 2011). In some cases, coup-proofing may even provoke coups (Harkness 2018; Sudduth 2017).

How else might leaders seek to protect themselves from a coup? An important option is increasing or decreasing the military’s political power. This approach to managing civil–military relations has low material costs, can be implemented quickly, and avoids some of the trade-offs of other strategies—yet entails substantial risks. Existing work focuses on power-sharing as a mechanism of credible commitment in autocracies (eg Meng, 2020; Boix and Svolik, 2013) and do not focus on the military-leader interaction specifically or look at overarching “strategies of control” in civil-military relations (Brooks and White, 2022) and do not examine how changes in civil–military power-sharing affect coups. This should, however, be a critical element in precipitating or insulating against coups—given that the vast majority of coups are carried out by members the military rather than civilian elites (Chin et al., 2021). There is historical evidence that the addition of military officers to government can protect against coups—for example, this was critical in insulating Zambian president Kenneth Kaunda from coups (Lindermann, 2011: 26). Yet, removal of military officers from their positions in government can seemingly precipitate a coup, as was the case in the 1984 coup attempt against Sevni Kountche in Niger (Decalo, 1990: 277-9).

Theory

I put forward a theory of how changing the military’s representation in government affects the probability of coup attempts. The theory applies in all democratic² and non-democratic regimes where the government and the military are distinct actors. This excludes “junta”-type regimes, where the military as an institution rules the state (Geddes et al., 2014).

The military can seek to resolve a dispute with the government by staging a coup. However, coups have uncertain chances of success—requiring significant coordination (Singh, 2014), and the potential costs of failure are severe. For this reason, coups are rare. Absent a coup, the leader and the military must still come to some agreement regarding their relative power. Within these agreements, there is still the potential for broad variation in civil–military relations (Feaver, 1999).

If it is possible for leaders and the military to come to agreement, why are there coups? Bargaining theory suggests two primary mechanisms: information problems and the inability to make credible commitments (Fearon, 1995; Powell, 2012). Both information problems (e.g., Svolik, 2012) and the inability to make credible commitments (e.g., Bell, 2016; Sudduth, 2017) have been put forward as major drivers of coups. I argue that these problems may be ameliorated when the leader adds military officers to the government and are exacerbated when they are removed. This approach to coup-proofing relates to the more general phenomenon of power-sharing in autocracies (eg Boix and Svolik, 2013; Meng, 2020)—whereby dictators surrender key positions and build institutions that constrain their power in order to reduce commitment problems However, the focus here is on power-sharing with military elites.

The military may launch a coup when it anticipates that the leader is attempting to take away power, autonomy, or resources. The addition of military officers gives the military a view into and a say over the policy-making process and constrains the leader’s power to impinge on these military prerogatives; it demonstrates for the military that a coup is not necessary to gain power or protect military interests. With this approach, the leader can give up some political power to the military in order to avoid losing all of it. Bringing officers into the government provides military oversight of the leader, reduces the leader’s control over the military, and provides it with influence over state policy. For example, the periodic appointments of military officers to high-level positions in the government helped Zambian president Kenneth Kaunda (r. 1964–1991) avoid coups. The former head of the army, General Malimba Masheke, stated that the appointment of military officers to high-level government positions by Kaunda showed “that to arrive in higher office, you do not need to take up arms” (Lindemann, 2011: 26). This theory conceives of the leader as willingly giving up positions to the military to make a credible commitment—yet critically, this occurs in the shadow of a coup. The objective in surrendering some power is to avoid losing it all in a coup.

In contrast, removing military officers from government can increase the likelihood of a coup. The removal of officers gives the military reduced oversight over the leader and less say in future policy-making, and it increases the leader’s power over the military and relevant areas of state policy. Absent the oversight, veto points, and power provided by actual seats in government, any promises that are made by the leader to the military carry little weight.

Exogenous factors may reduce the threat that the military poses and motivate the leader to remove its officers from positions of power. This might be a reduction in the military’s prestige after defeat in war, which allowed, for example, Nasser to marginalize the military in Egypt in 1967.³ However, I assume that even absent these exogenous shifts, leaders facing conflictual civil–military relations aim to increase their power over and autonomy from the military. Brooks and White (2022) conceive of civil–military relations as a balancing act for leaders facing politically active militaries. They balance the “governance imperative”—the desire to increase power over policy and resource allocation—with the “coup prevention imperative.” The value that leaders place on governance can undermine coup prevention.

This may have been what occurred, for example, in the October 1984 coup attempt against Seyni Kountche in Niger. Initially, the military was a major part of Kountche’s government; however, Kountche sought to appear less a military strongman and more a civilian ruler to domestic and international audiences. Accordingly, he began to remove military officers from the cabinet and replace them with civilians. This process of civilianization sparked widespread anger in the officer corps and led to the 1984 coup attempt (Decalo, 1990: 277-9).

H1: The removal of military officers from the government will increase the likelihood of a coup attempt while the addition of military officers reduces it.

Research design

I examine how changes in military participation in government (MPG) affect the likelihood of a coup. The theory’s expectations regarding changes in MPG require data that capture fine-grained variation both by country and year. I use White’s (2017) data on the number of active-duty military officers in national decision-making bodies in the executive branch (e.g., cabinets and state councils)—the Military Participation in Government (MPG) data. The MPG data covers cabinets, state councils, presidiums, and privy councils.

From the data, I generated ΔMPG. This is the net change in the number of seats in executive branch decision-making bodies that are held by active-duty military officers from the previous year to the present. This is intended to capture changes that the political leader has made to military representation in the leading decision-making bodies of government. The number can be negative (if there was a net decrease), positive (if there was an increase), or zero (if there was no change). Values for this variable range from −16 to 24. In the MPG data, 1,348 country-years out of 6376—or 21.14%—have values other than 0 for ΔMPG.

To measure coup risk, I generated a propensity score from a logistic regression that builds on Powell’s (2012) analysis of coups. The full specification of this model is detailed in the Appendix. From this, yearly predictions of coup attempt risk were generated for each country. This was included in all models. While ultimately the data used is observational and accordingly any statistical tests demonstrate associations rather than a causal relationship, the inclusion of the coup risk measure can serve to ameliorate concerns that any observed associations stem from leaders changing MPG in response to coup risk and not the impact of changes in MPG on coup propensity. Table 1 shows descriptive statistics for the variables included in the main analyses.

Table 1.

Descriptive statistics.

Variables	Mean	Standard Deviation	Minimum	Maximum	Observations
Coup risk	0.048	0.061	<0.001	0.399	5845
ΔMPG	0.017	1.320	−16	24	5567
MPG	1.264	2.439	0	24	5683
Government size	23.637	12.840	3	141	5683
ΔGovernment size	0.147	4.021	−42	53	5567
Coup attempt	0.044	0.206	0	1	7093

The dependent variable is coup attempt in a country-year, for which I use Powell and Thyne’s (2011) data. Here, the independent variable is ΔMPG. Since ΔMPG encompasses both increases and decreases together in a linear term, I also ran analyses disaggregating ΔMPG into two binary indicators for Any MPG Increases and Any MPG Decreases. For more nuanced analysis, I also disaggregated these binary indicators into additional indicators according to the scale of the change—that is, separate indicators for 1, 2, 3–5, and 6 or more added (or subtracted). The selection of these cut points was not arbitrary. Ideally, the model would include separate indicators for every level of change moving up and down by single units. However, changes in MPG of a greater absolute value are relatively rare such that separate coefficients could not be estimated at every level of ΔMPG. Instead, the aggregation of changes at 1, 2, 3–5, and 6 were chosen to allow for the estimation of coefficients and to capture non-linear relationships.

Additionally, I created another version of the independent variable which captures percent changes in MPG in terms of the overall size of the government in the prior year. For example, if two military officers were added to a cabinet of 20, the percentage change would be a 10% increase. Yet, if the same number were added to a cabinet of 30, the percentage change would be 6.667%. The inclusion of this measure offers a standardized way of measuring changes in MPG across regimes that may have cabinets of different sizes.

As additional controls, I included variables capturing extant levels of MPG, the overall size of the government, and the change in the overall size of the government to ensure that any effect observed for ΔMPG was not just capturing changes in the size of the overall government. Logistic regression was used given the binary dependent variable (coup attempt in a country-year).⁴ Robust standard errors were calculated by clustering on country and bootstrapped,⁵ given the use of the coup risk measures, which are predictions of the probability of a coup. All right-hand side variables were lagged by one-year to further ameliorate reverse causality. In line with the scope conditions—which require regimes in which the civilian political leadership and military are distinct actors, I excluded country-years where the regime-type in the Geddes, Wright and Frantz (2014) data was a military regime in the prior year. Table 2 shows the results for this analysis.

Table 2.

Changes in military involvement in government and coup attempts.

	ΔMPG)	% Change	Binary indicators	Multiple indicators
ΔMPG	−0.139**
	(0.044)
ΔMPG % change in gov.		−2.520**
		(0.827)
MPG decrease			0.571**
			(0.211)
MPG increase			0.571*
			(0.242)
Increase1				0.677*
				(0.283)
Increase2				0.804+
				(0.459)
Increase3–5				−0.539
				(0.709)
Increase>5				−1.104+
				(0.600)
Decrease1				0.444+
				(0.265)
Decrease2				0.478
				(0.448)
Decrease3–5				0.132
				(0.448)
Decrease>5				1.935**
				(0.418)
ΔGovernment	−0.005		−0.014	−0.007
	(0.020)		(0.019)	(0.020)
MPG	0.036	0.036	−0.001	0.017
	(0.033)	(0.029)	(0.035)	(0.034)
Gov. size	−0.004		−0.003	−0.005
	(0.010)		(0.010)	(0.010)
Coup risk	9.939**	9.981**	9.560**	9.528**
	(1.073)	(0.981)	(1.007)	(1.060)
Constant	−3.863**	−3.945**	−3.936**	−3.912**
	(0.264)	(0.141)	(0.262)	(0.268)
Observations	4884	4884	4884	4884

Robust, bootstrapped standard errors in parentheses.

**p < 0.01, *p < 0.05, + p < 0.1 (two-tailed).

The results present mixed support for the theory. While the sign on ΔMPG is negative and significant, as is the case with the models with ΔMPG measured in terms of percent of the government, the models with indicators for any net increase or decrease, respectively, are both positive and significant, suggesting that any change in MPG is associated with increased coup risk. However, the analysis that disaggregates MPG increases and decreases by magnitude (Model 3) provides a more nuanced view. This analysis suggests that small increases in MPG—that is, the addition of one officer—are associated with increased chances of a coup attempt. Large increases (the addition of six or more officers), however, are associated with a reduction in coup risk—in line with the theory’s expectations. In contrast, small decreases in military representation do not exert a significant impact on coup risk, while large decreases—the removal of six or more officers—are very strongly associated with an increased likelihood of a coup—also in line with the theory. The very large and significant coefficient for large net decreases in MPG may help to explain the overall negative coefficient on ΔMPG.

Figure 1 shows these effects, with the x-axis ranging (left-to-right) from large reductions to large increases in MPG. Relative to no change in MPG, large reductions (>5 officers) lead to a large increase in the chances of a coup attempt the following year—13.84% points,⁶ a more than four-fold increase in the probability of a coup from no change in MPG⁷—though there is a substantial uncertainty surrounding this prediction.⁸ In contrast, large increases reduce the chances of a coup attempt by 2.13% points (a reduction of 64.94% from the probability with no change in MPG). Both increases and decreases on the scale of one or two military officers generally increase the chances of coup, though outside of an increase of one, the effects do not approach conventional levels of statistical significance.

Figure 1.

Marginal effects of changes in MPG on probability of a coup attempt.

Additional analyses

I explored a number of different model specifications—found in the Online Appendix. These included the addition of measures of the levels of counter-balancing security organizations and military personnel in line with Powell (2012), and three additional specifications of the coup risk measure. In the first of these, I included a cubic polynomial of the time since the last coup attempt as a separate variable (e.g., Carter and Signorino 2010) from the coup risk measure. In the second, I changed the coup risk measure from one that measured the risk of a coup attempt to one that measured the risk of a successful coup. In the third, I reran the analyses with the individual components of the coup propensity score included in the analysis rather than the propensity score. In a fourth model, I replaced the coup risk measure with two-way fixed effects (country- and year-level). Across these analyses, the findings for the ΔMPG, ΔMPG as percentage of government, and binary indicator models are generally unchanged—with some slight reductions in significance.⁹ For the analysis of the multiple indicators, across the different alternative specifications there are losses of significance for the indicators that capture increases in MPG, but the overall pattern remains the same: small increases in MPG are associated with increased coup risk, larger increases with reduced coup risk. For MPG decreases, the effects are largely unchanged: large reductions in MPG are strongly (and significantly) associated with increases in the likelihood of a coup, while most smaller decreases exert a positive, but not statistically significant effect.

Conclusion and implications

The results provide mixed support for the theory. Contrary to the expectation that increases in MPG should broadly insulate leaders against coups, I find evidence that this only applies to large increases, while smaller-scale increases are actually associated with increased coup risk. This might occur because the coups that come after small increases in military participation in government reflect cases where leaders underestimate the minimum offer of political power that the military will accept before resorting to a coup and make an offer less than that.

It is clear, however, that large decreases in MPG substantially increase coup risk. A key question, then, is why leaders would seek such a dramatic shift, given the risk of a coup? In addition to the desire of leaders to want to increase their share of political power, an additional factor can be information problems. Where the leader has poor information regarding military preferences or coup propensity, they may overstep in removing military officers from government. Even if the act of removing the officers does not provoke a coup, their absence may make one more likely, as the leader loses a potentially important channel to and from the military in the form of government ministers in uniform. Accordingly, the leader may be more likely to miscalculate in future interactions with their military.

The results presented here demonstrate important statistical associations between changes in MPG and coup propensity. Critically, these results are suggestive and do not demonstrate a causal connection. The inability to randomly assign changes in MPG to governments hinders identifying a causal relationship between these changes and coup propensity. Future work might build on the analysis here and combine multiple empirical approaches, such as process tracing, difference-in-difference analysis stemming from leadership changes, surveys of military elites in different states (where possible), or experimental work in order to gain further leverage over the question of how changes in military political power affect coup propensity.

While suggestive, the results presented here do support the claim that more nuance is required when evaluating the desirability of civilianization of politics in situations where the military already holds political power. Civilian control of the military is a vital and necessary component of democracy and of democratic consolidation, and this article does not argue that emerging democracies should avoid civilianization of politics or establishing civilian control of the military. Rather, the results presented here suggest that civilization of politics should proceed carefully. When moving from conditions of high-levels of military involvement in government, small-scale reductions over time are relatively low-risk and are desirable from a normative perspective; however, sharp and sudden reductions can exacerbate government-military tensions provoke coup attempts by the military, undoing efforts at civilianization and potentially undoing hard-won gains in democratization.

Supplemental Material

Supplemental Material - Getting a Seat at the Table: Changes in Military Participation in Government and Coups

Supplemental Material for Getting a Seat at the Table: Changes in Military Participation in Government and Coups by Peter B White in Research & Politics

Footnotes

Acknowledgements

The author is grateful to Risa Brooks, Michael Kenwick, Anne Meng, and the editors and anonymous reviewers at Research and Politics for their valuable feedback. An earlier version of this article was presented at the 2018 meeting of the International Studies Association Annual Conference. The data used in this article were collected with support from a National Science Foundation Dissertation Improvement Grant (SES #1424001). This article does not necessarily reflect the views of the National Science Foundation or the United States government.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Science Foundation (SES #1424001).

Correction (June 2025):

The article has been updated with correct dataverse link in the supplementary material section. For more details, please see the correction notice: .

ORCID iD

Peter B White

Supplemental Material

Supplemental material for this article is available online.

The files can be found at

Notes

References

Albrecht

Eibl

(2018) How to keep officers in the barracks: causes, agents, and types of military coups. International Studies Quarterly 62(2): 315–328.

Belkin

Schofer

(2003) Toward a structural understanding of coup risk. Journal of Conflict Resolution 47(5): 594–620.

Bell

(2016) Coup D'etat and democracy. Comparative Political Studies 49(9): 1167–1200.

Boix

Carles

Svolik

Milan

(2013) The foundations of limited authoritarian government: Institutions, commitment and power-sharing in dictatorships. The Journal of Politics 75(2): 300–316.

Brooks

Risa

(2008) Shaping Strategy: The Civil-military Politics of Strategic Assessment. Princeton University Press.

Brooks

White

(2022) Oust the leader, keep the regime? Autocratic civil-military relations and coup behavior in the Tunisian and Egyptian militaries during the 2011 arab spring. Security Studies 31(1): 118–151.

Carter

Signorino

(2010) Back to the future: modeling time dependence in binary data. Political Analysis 18(3): 271–292.

Chin

John J

Carter

David B

Wright

Joseph G

(2021) The varieties of Coups D’etata: Introducing the Colpus Dataset. International Studies Quarterly 65(4): 1040–1051.

Collier

Hoeffler

(2005) Coup Traps: Why Does Africa Have So Many Coups D'état? Oxford: Centre for the Study of African Economies, Department of Economics, University of Oxford.

10.

De Bruin

Erica

(2020) How to Prevent Coups d’Etat: Counterbalancing and Regime Survival. Cornell University Press.

11.

Decalo

(1990) Coups and Army Rule in Africa. New Haven: Yale University Press.

12.

Fearon

(1995) Rationalist explanations for war. International Organization 49(3): 379–414.

13.

Feaver

(1999) Civil-military relations. Annual Review of Political Science 2(1): 211–241.

14.

Finer Edward

(1973) The Man on Horseback: The Role of the Military in Politics. New Brunswick, NJ: Transaction Publishers.

15.

Geddes

(1999) What do we know about democratization after twenty years? Annual Review of Political Science 2(1): 115–144.

16.

Geddes

Wright

Frantz

(2014) Autocratic breakdown and regime transitions: a new data set. Perspectives on Politics 12(2): 313–331.

17.

Hanmer

Ozan Kalkan.

(2013) Behind the curve: clarifying the best approach to calculating predicted probabilities and marginal effects from limited dependent variable models. American Journal of Political Science 57(1): 263–277.

18.

Harkness

(2018) When Soldiers Rebel: Ethnic Armies and Political Instability in Africa. Cambridge: Cambridge University Press.

19.

Lindemann

(2011) The ethnic politics of coup avoidance: evidence from Zambia and Uganda. Africa Spectrum 46(2): 3–41.

20.

Meng

(2020) Constraining Dictatorship: From personalized rule to institutionalized regimes. Cambridge University Press.

21.

Piplani

Talmadge

(2016) When war helps civil-military relations: prolonged interstate conflict and the reduced risk of coups. Journal of Conflict Resolution 60(8): 1368–1394.

22.

Poe

Neal Tate

Keith

(1999) Repression of the human right to personal integrity revisited. International Studies Quarterly 43(2): 291–313.

23.

Powell

Thyne

(2011) Global instances of coups from 1950 to 2010: a new dataset. Journal of Peace Research 48(2): 249–259.

24.

Powell

(2012) Determinants of the attempting and outcome of coups d’état. Journal of Conflict Resolution 56(6): 1017–1040.

25.

Quinlivan

(1999) Coup-proofing: its practice and consequences in the Middle East. International Security 24(2): 131–165.

26.

Roessler

(2011) The enemy within: personal rule, coups, and civil war in africa. World Politics 63(2): 300–346.

27.

Singh

(2014) Seizing Power: The Strategic Logic of Military Coups. Baltimore, MA: Johns Hopkins University Press.

28.

Sudduth

(2017) Coup risk, coup-proofing, and leader survival. Journal of Peace Research 54(1): 3–15.

29.

Svolik

(2012) The Politics of Authoritarian Rule. Cambridge: Cambridge University Press.

30.

Talmadge

(2015) The Dictator’s Army: Battlefield Effectivenesss in Authoritarian Regimes. Cornell University Press.

31.

Weeks

(2012) Strongmen and straw men: authoritarian regimes and the initiation of international conflict. American Political Science Review 106(2): 326–347.

32.

White

(2017) Crises and crisis generations: the long-term impact of international crises on military political participation. Security Studies 26(4): 575–605.

33.

White

(2020) The perils of peace: civil war peace agreements and military coups. The Journal of Politics 82(1): 104–118.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.29 MB