https://orcid.org/0000-0003-4930-2862
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
In this article, we assess the variation in civilian governments’ defense spending as a function of civil-military relations. We present a novel explanation based on the military’s presence in the political decision-making apparatus. We argue that the appointment of an active military officer in a key governmental position allows the government to make a more credible commitment to provide the military with adequate rents and thereby stabilizes civil-military relations. The appointment helps solve the commitment problem on the part of the government by reducing the coordination costs for the military to challenge the government more successfully in the event of the government’s defection. Hence, the military’s incentive to intervene in politics to prevent the government’s monopoly over rents lessens markedly. Defense spending increases as a result because this arrangement requires the government to honor its promises to distribute increased rents among the military members of the winning coalition. Our theory predicts higher defense spending in cases where the military has a presence in the top-level policy-making positions within a civilian government. We find strong empirical evidence in support of our argument using data from 1984 to 2011. Civilian governments with a military presence in the cabinet’s security-related portfolios experience up to a 20% increase in their defense spending on average. This study offers important insights into governments’ motivation to appoint active-duty military officers in key policy-making positions and the effects of such appointments on civil-military resource allocation.
En este artículo, evaluamos la variación del gasto en defensa de los gobiernos civiles en función de las relaciones cívico-militares. Presentamos una explicación novedosa basada en la presencia militar en el aparato de toma de decisiones políticas. Argumentamos que el nombramiento de un militar en activo en un puesto gubernamental clave permite al gobierno comprometerse de forma más creíble a proporcionar a los militares rentas adecuadas y, por tanto, permite estabilizar las relaciones cívico-militares. El nombramiento ayuda a mitigar el problema del compromiso por parte del gobierno al reducir los costes de coordinación de los militares para desafiar al gobierno con un mayor éxito en caso de deserción del gobierno. Por consiguiente, el incentivo de los militares para intervenir en la política con el fin de impedir los monopolios gubernamentales sobre las rentas disminuye de forma considerable. El gasto en defensa aumenta en consecuencia porque este acuerdo exige que el gobierno cumpla sus promesas de distribuir mayores rentas entre los miembros militares de la coalición ganadora. Nuestra teoría predice un mayor gasto en defensa en los casos en los que los militares están presentes en los puestos políticos de más alto nivel dentro de un gobierno civil. Encontramos sólidas pruebas empíricas en apoyo de nuestro argumento utilizando datos de 1984 a 2011. Los gobiernos civiles con presencia militar en los puestos relacionados con la seguridad del gabinete registran un aumento medio de hasta el 20% en sus gastos de defensa. Este estudio ofrece importantes perspectivas sobre la motivación de los gobiernos para nombrar a militares en activo para puestos clave en la formulación de políticas y los efectos de dichos nombramientos en la asignación de recursos cívico-militares.
Dans cet article, nous établissons un lien entre les dépenses de défense des gouvernements civils et les relations entre civils et militaires. Nous proposons une nouvelle explication fondée sur la présence militaire dans le mécanisme de prise de décision politique. Selon nous, quand un gouvernement nomme un officier en service à un rôle clé au sein du gouvernement, il renforce la crédibilité de son engagement à attribuer au corps militaire des rentes adéquates et donc, stabilise les relations entre civils et militaires. Grâce à cette nomination, le gouvernement lève tout doute concernant son engagement : ses coûts de coordination réduits, le corps militaire pourra plus facilement défier le gouvernement en cas de défection. Les forces armées seront d’autant moins tentées d’intervenir politiquement contre les monopoles du gouvernement sur les rentes. Par conséquent, les dépenses de défense augmentent : le gouvernement doit honorer ses promesses de distribution de davantage de rentes aux officiers de la coalition gagnante. Selon notre théorie, les dépenses de défense augmenteraient quand des officiers occupent des fonctions importantes dans la prise de décisions au sein d’un gouvernement civil. À l’aide de données datant de 1984 à 2011, nous obtenons des preuves empiriques convaincantes pour étayer notre propos. En cas de présence militaire dans les dossiers du conseil des ministres relatifs à la sécurité, les dépenses de défense des gouvernements civils augmentent en moyenne de 20 pour cent. Cette étude permet de mieux comprendre ce qui motive les gouvernements quand ils nomment des officiers en service à des fonctions clés du processus de prise de décisions. Elle montre par ailleurs les effets de ces nominations sur la répartition des ressources entre civils et militaires.
Acknowledgments
The authors thank Scott Wolford, Terrance Chapman, Patrick McDonald, Toby Rider, Tse-Min Lin, anonymous reviewers, and the editor.
Notes
1 For Bove and Nistico (Citation2014b), the period of the analysis is confined to 1984–2009. We have two additional years of observations. Military regime data that we use to draw the sample of civilian countries in our analysis is not available for the period after 2011. For some key variables, data are not available for recent years too.
2 However, they treat the hybrid categories of party-military and party-personal-military, as party-based regimes. These categories include governments that are formed by a military junta but transitioned to a party-based one because the leaders formed a political party to run the government. While this can be problematic given that the civilian governments of these categories are subject to some military influence, the categorization nonetheless serves our purpose. By drawing a line between the civilian government and the military, we seek to differentiate the incentives of the government to bargain with the military as having distinct organizational priorities. We, therefore, include party-military and party-personal-military hybrid categories in our sample of civilian governments, while excluding the military-personal hybrid category. We have explored some alternatives to GWF. The other available data by Cheibub, Gandhi, and Vreeland (Citation2010) treats even a political-party-based democratic government led by a former military officer as a military government, which is even more problematic in terms of modeling civil-military relations. Former military officials, although can be somewhat prone to side with the military, do not represent the military-organizational interests in their formal role. They also vary enormously, some having just a minimal experience of being conscripted at a young age.
3 We use the Hausman test to choose whether we need to run the fixed effects model or the random effects model (see Supplementary Table A2 in the Appendices). The test indicates that the estimates from random effects are not consistent. We also add the year-fixed effects because the coefficients for the temporal units are jointly significant.
4 We choose the order of autoregressive lags based on the theoretical justification discussed in the results section. However, for robustness check, we estimate additional models with up to (3,1) lags and the results show that the choice of ADL(1,1) is justified on empirical grounds as well.
5 Since we have a short time series, we assume that the variables are stationary. We have 27-time units, so the Nickell Bias is not a problem for our models with LDY.
6 We use annual defense spending as the main measure because of the standard practice in studies that explain variation in defense budgets. Following them, we controlled for different measures of state capacity (GDP, Population size) as well as the demand for defense budget (the size of the military), to address any confounding variations. That allows making inferences without transforming the variable itself as the proportion of GDP or government spending or standardizing the size of the military.
7 For robustness check that accounts for more intense international threat environment, we use: (i) a measure of more severe interstate conflict, MIDs that resulted in more than 100 battle-deaths, (ii) the number of conflicts present in the international system, and iii) the number of neighboring states (calculated based on the Direct Contiguity dataset of the Correlates of War Project [Stinnett et al. Citation2002]) as additional control variables. The last two variables indicate the number of potential fronts that a country might face in terms of its territorial rivalries and in other forms of international disputes.
8 In the robustness check models, we also add a log-transformed measure of the total population to control for the size of the country. Data also come from the World Bank’s World Development Indicators (World Bank Group Citation2016).