Defense Sector Politics

Ownership and control of defense industrial firms affords the military power, autonomy, and a claim to economic rents. Why do some countries succeed at shifting some or all such firms from military to civilian ownership and control, while others do not? I argue that differing configurations of relative civilian and military coalitional and institutional strength contribute to outcomes. Civilian and military must find party and executive-branch allies who can compete for them to craft or defeat legislation affecting their interests, but coalition building alone is insufficient. Actors must also cultivate executive bureaucratic institutions able to design and implement policies promoting their interests. How coalitions and institutions form, ossify, or fail to develop is assessed through a comparative study of Chile, Argentina (1983–1989), and Argentina (1989–1997).


Introduction
Ownership and control of defense industrial firms affords the military power, autonomy, and a claim to economic rents. That permits the military to favor some civilian economic concerns by investing in them and squeeze out others by competing with them and, indeed, prevent investment and exchange in more open and competitive markets. In government corridors, it translates into privileges, protections, and access to inside information; influence during the budgetary and procurement processes, and veto power over the projects of more dynamic and innovative civilian sectors. The secretive firms can foster corruption by military personnel and sew mistrust between inter-state rivals, who lack information about the off-budget revenue the firms generate. Such revenue also bedevils the civilian authorities, whose leverage vis-à-vis the military is diluted by the latter's access to sources of finance outside the budget.
Defense industrial firms and political efforts to reform them by transferring them to civilian ownership and control are contemporary, global issues. The conflict over such reform or its failure is a critical distributional struggle in countries where such sectors often perform an outsized role as well as constitute the key source of the military's power in its relationship with state and society. In Argentina before reform in the 1990s, the military owned in whole or in part 45 industrial and service-sector firms that employed 33,000 people directly; enjoyed a monopoly or near monopoly in the mining, weapons, steel, petrochemical, and naval sectors; and, at their peak, accounted for a third of gross domestic product (GDP). In Chile, five decades of neoliberal policy notwithstanding, the same three firms remain in military hands and wielders of a de facto monopoly on contracts to service arms imported with lavish off-budget copper export revenue.
The political economy of defense industrialization and of civil-military relations increasingly preoccupies scholars. After countries defied the predictions of globalization theorists and continued pursuing defense industrialization well into the post-Cold War era, scholars began debating its drivers, then turned to the inner workings of the policy making process (Kurç and Neuman 2017). In parallel, Alfred Stepan told scholars that "if there is a major debate in the country about the size, direction, content, and control of a domestic military-industrial complex, then it becomes an issue that requires separate analytic and political attention" (1988: 82). Successors responded with studies of military enterprise development (Mani 2007), military export levy reform (Grimes and Pion-Berlin 2019), and levels of military production (Acuña and Smith 1995). Yet military industrial reform escaped systematic, comparative inquiry. Why do leaders who craft policies transferring some or all defense industrial firms from military to public or private civilian ownership and control either succeed or fail at what they set out to do?
This article argues that differing configurations of relative civilian and military coalitional and institutional strength contribute to divergent outcomes. Civilian and military must surmount diverse obstacles to translate their interests into policy. Actors must find party and executive-branch allies who can compete on their behalf to craft or defeat legislation affecting their interests, but coalition building alone is not enough. Actors must also cultivate executive bureaucratic institutions possessing the authority, information, and expertise to design and impose policies promoting their interests. Coalition formation and institution creation become critical. Where civilian coalitions and institutions emerge stronger, all or most firms will be transferred; where military coalitions and institutions lead the way, reform will be stalled. The sector will be partly reformed where military coalitions and civilian institutions both prevail.
Which side emerges with relatively stronger coalitions and institutions depends on certain factors. Parties and executive-branch actors drawn to the military must derive some political or electoral advantage (or at least no disadvantage) from supporting economic institutions that dilute governmental leverage. Then, they must obtain the requisite support to introduce reform and prevail in the congress, building coalitions to overcome party system fragmentation as needed. Executive bureaucratic institutions whom civilian or military would enlist to craft or defeat reform must be obligated to do so. They then must locate the requisite authority, information, and technical capacity to shape the relationship between military and society. These battles will be assessed through two comparative studies of Chile (1990-present), Argentina (1983Argentina ( -1989, and Argentina (1989Argentina ( -1997, followed by analysis of three alternative explanations.

Defense Industrial Policy Making
Scholars debate the drivers of defense industrial policy making, but they overlook the politics of civil-military rivalry. One group argues that the forces of globalization lead states to privatize their arms industries (Hayward 2000;Struys 2004). That imposes losses on powerful actors, including the military, workers, partners, and suppliers, who may align to oppose it. If one is going to craft reform to attract investment, then they must be able to get by the opposition first. A rival group holds that a quest for power, wealth, and prestige motivates states to avoid change and, instead, pursue greater industrialization (Evans 1986;Bitzinger 2003). Yet reform would create gains for the winners, including private actors and multinational corporations, so the status quo's defenders must be able to stop the coalition for change. What we need in order to determine why countries go different ways is a theory of the power relations that shape whether and how reform emerges. This article argues that the relative coalitional and institutional strength of civilians and the military can explain the occurrence (or absence) of defense industrial reform.

Civilian-Military Power Relations
Evidence of a role for coalitional strength in legislative battles between politicians and the military is mounting (Negretto 2013;Grimes and Pion-Berlin 2019). The military's influence dissipates without the support of actors with sufficient legislative and institutional representation to shape outcomes, including the president or prime minister, defense portfolio, and a congressional majority. Nor can the civilian take her power for granted because her own erstwhile supporters may align with the military. In the contest to shape legislative defense industrial reform outcomes in their favor, reform-minded civilian leaders and the militaries who oppose them articulate clear and well-defined preferences, and the defense industry creates its own bases of support due to links to other industries and to labor and private stakeholders' investments in the sector (Réal-Pinto 2017). Thus, civilian and military actors would be inclined to build or elicit the support of coalitions to craft or defeat reform legislatively. For the military, beholden as it is to the executive branch, this means being discrete and aiming for parties, the executive, and/or the defense portfolio to adopt positions that converge with its own.
Yet coalitional strength only gets an actor so far when the distributive conflict enters the executive bureaucracy. The defense industrial bureaucracy may be structured to facilitate military power. First, it enhances the military presence if civilians do not occupy the top post in the defense ministry or its staff positions, a defense ministry does not exist, the post of defense minister is reserved for the top-ranking officer, or there is a tradition of selecting an officer. Even if the defense portfolio maximizes the civilian presence, command authority may not belong to that minister but to the commanders of the service branches. Second, the defense ministry may not be empowered to make defense industrial decisions, reserved as they are for the service branch ministry or other military agency administering the firms. Third, coalitions must grant these bureaucracies wider mandates due to the complexity of defense industrial reform and the limits to information and transparency in the sector. This means that de facto or de jure military agencies will get to choose their program and define their goals and procedures. Together, these features define an officialdom in which decisions are reserved for military officers. Changes desired and legislated by the coalition that favors reform would neither be ordered nor supported, as agency heads would use their enhanced agenda control to perpetuate the status quo.
To surmount these obstacles, a civilian leader must come to power and build a dominant coalition for reform and then institutionalize the coalition's strength by reforming the executive bureaucracy from the top down. A president or prime minister who is pro-reform and leads a winning coalition must make political appointments in the defense ministry, beginning with that of a like-minded defense minister (Serra 2010). The government creates a single department for crafting defense industrial policy that is controlled by the executive through the command of the defense ministry. This agency gradually acquires a permanent nucleus of disciplined civilian staff who replace military staff and civilian contractors (Bruneau and Tollefson 2006). It begins to grow, build expertise, assume more authority, and, ultimately, serve the cause of reforming the defense industry. In general, unless civilian coalitions institutionalize their power first, military influence in the bureaucracy will prevent even the strongest coalitions from realizing reform.
By contrast, even the strongest military coalitions cannot stop equally dominant civilian institutions from pursuing reform without the participation of the legislature. A leader thwarted legislatively may cite legal powers of the executive to create new tools for acting unilaterally. And although this typically limits the civilian leader to pursuing partial reforms, she may impose serious losses nevertheless if she commands an institution affording her a monopoly on authority, information, and expertise. Accordingly, the military must maintain the requisite support to block any effort to reform military-dominated bureaucracies legislatively.

Coalitions and Institution Building
Executive bureaucratic institutions are an essential structural support for any coalition intent on executing or preventing sectoral reform (Nelson and Waterbury 1989). Dispersing authority across agencies can weaken them, duplicate functions, and undermine decision making. Concentrating formal decision-making capacity in its own hands by creating and cultivating a special-purpose executive bureaucratic institution helps a coalition to mitigate these problems. By altering key patterns of capacity and obligation, institution creation produces a setting in which a coalition retains control over the design and execution of its policy goals (Moe 2005).
Not all institutions are equally capable of realizing the goal of defense industrial reform. Creating a capable institution requires that the central administration created to set defense industrial policy controls credit and can intervene in sectors by appointing firm directors and controlling enterprise reforms (Evans et al. 1985). Control of credit is needed to finance deficits that may arise during the transition and block its completion (e.g., from indemnifying workers or selling indebted firms without their liabilities). Power to intervene in sectors matters because enterprise reform must be kept separate from uniformed enterprise management, who have no incentive to self-reform and will resist it, and because officials must be able to amend earlier proposals to craft changes that would gain political acceptability support at the firm level.

Method and Cases
This study of defense industrial reform uses process tracing within two most-similar systems designs. Upon comparing cases of defense industrial reform according to the variables of relative civilian and military coalitional and institutional strength, I selected three cases whose values differed sharply. In Chile (1990-present), dominant military coalitions and military executive bureaucratic institutions emerged, and reform failed. In Argentina (1983Argentina ( -1989, dominant military coalitions and civilian executive bureaucratic institutions formed, and reform partially succeeded. In Argentina (1989Argentina ( -1997, dominant civilian coalitions and civilian institutions arose, and reform succeeded. Argentina and Chile's defense industrial firms were autonomous, evolved from military infrastructure, and were created amid arms embargoes imposed by advanced powers. In addition to controlling for these variables, the comparative analysis of these cases controls for alternative explanations by holding constant the variables of authoritarian legacy, economic development, diplomatic initiatives, and strategic environment.
First, both countries democratized after long bouts with military rule during which the militaries permeated government, meaning their militaries would have an equally large stake in protecting their interests. Second, both were upper-middleincome countries with above average levels of development, as measured by GDP per capita, employment rates, wage levels, strong middle classes, and lower levels of poverty. Such economic successes would have provided equally sound platforms for democratic reforms. Third, diplomatic initiatives between Argentina and Chile (including the settling of border disputes and forging of nuclear pacts) ended rivalry, eliminating existential threats which might have led politicians to prioritize defense. And fourth, both reside in a low-threat environment, meaning it would have been harder for politicians to prioritize defense issues. The first comparison is cross-national: Argentina and Chile; the second is cross-temporal and intra-national: Argentina (1983Argentina ( -1989 and Argentina (1989Argentina ( -1997.
I used within-case process tracing to determine if differences in relative coalitional and institutional strength actually shaped outcomes, identifying a continuous chain of events for each case and analyzing processes, sequences, and conjunctures of events to make inferences about causal explanations of that case. My comparisons are informed by diverse types of evidence. I drew, first, on local and external analyses of the defense industrial politics in each of these countries and periods, government documents and legislation, and intelligence from the World Bank. I especially drew on 400 newspaper articles and texts of legislative bills obtained in the archives of the University of Buenos Aires's Institute of Economic Research. I also relied upon 32 in-depth interviews that I conducted in 2018 in Santiago, Chile, and Buenos Aires, Argentina. My interviewees were academics who had analyzed the defense industries and current and former defense ministers, agency heads, party members, and policy advisors. I analyzed the sources using the method of triangulation, whereby data is cross-checked from multiple sources.

Chile
In Chile, the page turned on military rule in 1990, but the military remained a player in the economy in 2021. Each service branch operated its own legally autonomous force-oriented business. FAMAE was the Army's munitions and maintenance firm, ASMAR the Navy's naval construction and repair firm, and ENAER the Air Force's aircraft manufacturer and repair firm. More formidable still was the dictatorship's political legacy, particularly a constitution granting the military non-elected senators and control of a security council that could veto the president's decisions. In September 2005, the Concertación and Alianza coalitions of parties allied to approve constitutional changes stripping the military of these and other tutelary powers. One by one, governments of the Left and Right sought to seize the military's firms to hand them to civilians next. Their efforts failed, such that "the one area where the military does retain considerable autonomy is in defense production" (Pion-Berlin and Martínez 2017: 102).
The fight for the defense industry was one theater of the conflict over the modernization of the state. The corruption allegations involving bureaucrats, state firms, and dubious purchases that surfaced with increasing regularity beginning in the early 2000s also engulfed defense. In 2018, falsified invoices for US$200 million in alleged sales of weapons linked to FAMAE arose in connection with the investigation into Milicogate, a giant corruption scandal first unearthed in 2011 involving military officers, copper revenues, and payments for services never rendered. Besides tanking confidence in government, the quality of democracy, and anti-corruption measures, the allegations launched pro-transparency initiatives. Militaryrun firms were targeted because they were viewed as uniquely conducive to corruption, and for good reason. Immune from the reporting requirements of the Sistema de Empresas (Public Enterprise System-SEP), the only parastatals which did not officially answer to a minister or any other government official, and wielders of a monopoly on contracts to service arms imported with rich off-budget copper export revenue, the military firms were flush with wealth and yet operated in darkness.

Military Coalitional Strength
In 2007, Partido Socialista (Socialist Party-PS) President Michelle Bachelet (2006)(2007)(2008)(2009)(2010) vowed to send Congress a bill forcing military−run firms to integrate two directors appointed by the Executive and adopt accounting and financial standards like those of other parastatals. Word got out about the military's opposition and that of the ENAER and ASMAR unions, the latter expecting reform to precipitate privatization and reduce wages, jobs, and the sources of their labor power. Many ENAER union leaders, company staff, and workers were former or retired military or their children: "in short, the company's unions are made up of people who listen to the suggestions of the high command." 1 Led by PS and Partido Demócrata Cristiano (Christian Democrat Party-PDC) militants, so, too, did they form part of the electoral base of the Concertación coalition of parties, which ruled the Chamber of Deputies. Unable to take a position that would alienate its union support, the coalition sided with labor and the military. 2 The military's strength was poised to exhibit itself in Congress, resulting in legislative defeat.
The military party coalition delayed the presentation of the initiative, preventing Bachelet from submitting it by April 2007 to the Lower Chamber. It would now be sent in December 2007, said Finance Minister Andrés Velasco, but the coalition's wrangling again kept the government from meeting its deadline. When the Executive finally dispatched the bill to the Chamber Finance Committee, in March 2008, it was promptly torpedoed by the committee's Concertación majority. 3 Thus, defense industrial corporate governance reform failed when the coalition aligned with the military in opposition to reform became strong enough to resist its imposition legislatively by the coalition that favored it.

Military Dominance over a Weakened Ministry of Defense
Frustrated by Congress, the Executive sought to begin reform without it. In May 2012, Defense Minister Andrés Allamand of the Renovación Nacional (National Renewal-RN) President Sebastian Piñera Administration (2010-2014) announced a partnership negotiated by his portfolio between ENAER and European defense firm Airbus Military. The Air Force would remain the manager in the firm's repair and maintenance divisions and the Finance Ministry would remain the firm's owner, but the Air Force would otherwise be subordinated to Airbus Military commercial and industrial management in a relationship like that of a subcontractor. Military objections and whispers of intent to subvert the reform quickly surfaced: the firm's helms "include people drawn from the Air Force who, along with receiving a salary in ENAER, are receiving a pension and who, therefore, will not stand idly by," said one source. 4 The policy also caused a stir because it demanded much from a relatively weak civilian bureaucracy, not to mention the cooperation of military officers in key positions of power. The MoND's Department of Defense Industry could merely audit the military−run firms, and, in any case, remained under the control of military officers (Flisfisch and Robledo 2012). Partido por la Democracia (Party for Democracy-PPD) President Ricardo Lagos (2000-06) had proposed to create an institution for the sector empowered to name directors and intervene in firms, but the military thwarted his plans by backing a weaker alternative favored by the party Right. 5 Now, in 2012, many doubted that the MoND could execute its program. For his part, Allamand pointed to a study, commissioned in 2009 by then Defense Minister José Goñi, which concluded that integrating private capital and management into military−run firms was possible and legal. 6 But few readings of the informal context could have been so sanguine. Luis Hernán Paul Fresno, the outside advisor whom Allamand was forced to hire to negotiate with the firms' officer−directors, was meeting with military resistance. 7 The bureaucracy was set to become a staging ground for the reappearance of the military's political influence, ensuring the perpetuation of the status quo.
When Chamber Defense Committee members Deputies Ricardo Rincon (DC) and Tucapel Jiménez (PPD) persuaded the Comptroller to review the legality of the ENAER−Airbus deal, the Comptroller ruled that the MoND lacked the requisite authority. Yet the officers in the de facto military bureaucracy did not support Allamand's policies even before the ruling. Military inaction tempered bureaucratic advances, making the reform drive more vulnerable to legal challenges than it would have been had the bureaucracy been firmly in civilian hands. As Rincon put it, "here there is an absolute lack of public policy for the sector by the Ministry of Defense and the government. Even with a press of three ministers, it goes nowhere." 8 In short, the incorporation of private capital into military firms failed when the military in the bureaucracy emerged strong enough to resist its imposition administratively by the civilians who favored it.

Military Coalitional Strength
Though the military objected privately to attacks on its interests, its opposition always surfaced. When the second Socialist Michelle Bachelet Administration (2014-2018) said, in October 2014, that it was finalizing a bill to permit the 5 Miguel Navarro Mesa, Professor at the National Academy of Political and Strategic Studies, interview with the author, May 8, 2018, Santiago, Chile. 6 El Periodista, October 15, 2012. 7 Luis Hernán Paul Fresno, interview with the author, July 12, 2018, Santiago, Chile. 8 Cambio21, "Enaer: El forzoso aterrizaje de una privatización encubierta," September 21, 2013. appointment of majorities of civilians to the boards of military−run firms, newspapers reported on the military's complaints soon after: "the receptivity of the military [to reform]? They continue to show reluctance," scribbled one. 9 Defense Minister Jorge Burgos (PDC) pushed hard for the bill nevertheless, as did the party Left and Right in Congress. 10 Milicogate had put the military on the defensive and rendered reform electorally profitable. And the ASMAR unions, seeing reform as a path to victory in their fight with the firm's management, whom had long refused to grant the unions the right to collectively bargain, had backed the plan, splitting the defense sector's unions and making it possible for Concertación to support reform. It looked like the civilian coalition would sail to victory.
Yet the coalitions began to change again. By May 2015, the civilian coalition was in crisis. Bachelet's government had also pursued tax, educational, constitutional, labor market, and pension reforms, but many of these began to fail in design or implementation, prompting widespread opposition to her rule. A slowing economy catalyzed the opposition, as did allegations that an ally, Interior Minister Rodrigo Penailillo, was tied to a firm mired in a campaign finance scandal. For months, the military had been warning of conflicts it said would result from mixing civilians and military in firm directories. Viewing herself to be unprepared for a long fight with the military, Bachelet withdrew the defense portfolio from the civilian coalition by sacking Penailillo and shifting Burgos from Defense to Interior, then appointing former Radical Party Senator and Senate Defense Committee member José Antonio Gómez as Defense Minister. An established politician closely tied to the unions, who had strongly opposed Allamand's ENAER reforms in a Senate hearing, 11 Gómez "represented no threat to the military, and the military knew it," 12 said Jaime Baeza Freer, Deputy Director of the National Academy of Political and Strategic Studies (ANEPE), a graduate school administered by the MoND, between 2014 and 2017. The Executive and defense portfolio, now legs of the military coalition, were set to in effect allow the military into government corridors to exert its influence, ensuring defeat.
The military executive coalition said little about the bill for a year and a half. Only in late 2016 did MoND quietly dispatch the project to the Ministry General Secretariat of the Presidency (SEGPRES), its last stop before Congress (Gobierno de Chile et al. 2017: 138). But SEGPRES then went silent. Deputy Marcelo Schilling (PS) and House Defense Committee leaders resorted to using Defense Minister Gómez's presence in their chambers to inquire about the bill's whereabouts. It had left his orbit, Gómez insisted. 13 But the pro−reform lawmakers could see the writing on the wall: despite agreement in Congress, the "political will to advance 9 Economía y Negocios, "Defensa afina proyecto que moderniza directorios de la industria militar," October 19, 2014. 10 Jorge Burgos, personal correspondence with the author, July 24, 2018. 11 República de Chile, Senado, "Realizan rondas de audiencias para conocer situacion de la Empresa Nacional de Aeronáutica," June 14, 2012, senado.cl. 12 Interview with the author, April 24, 2018, Santiago, Chile. 13 La Segunda, "Pelea de ministros: El "raspacachos" de Gómez a Eyzaguirre," August 11, 2017.
[in the Executive] was lacking," as Deputy Jaime Pilowsky (PDC) lamented. 14 The defense industrial reform bill became "one of many reforms killed during Gómez's tenure" said Freer. 15 Jorge Burgos, Gómez's predecessor and the chief proponent of the reform, agreed. 16 The reform bill's prospects have not improved as of this writing. In March 2018, Defense Minister Alberto Espina of the second Sebastian Piñera Administration (RN) (2018-present) told the inquiring Senate and House Defense Committees that the legislation was "not among the priorities the SEGPRES was handling." 17 Asked why, Miguel Navarro Mesa, a professor at ANEPE, said that Piñera had ordered Minister Espina "to maintain very good relations with the military," whose contestation of reform had apparently persuaded Piñera, too, that resultant civil−military conflict within defense industrial firms would erase the potential efficiency gains from reform. 18 Thus, in 2015-2016 and in 2018, the drive for defense industrial corporate governance reform stalled when the coalition aligned with the military in opposition to reform emerged strong enough to prevent its introduction legislatively by the coalition that favored it.

Argentina, 1983-1989
In land, labor, and capital, the Argentine military was the biggest player in the economy at the dawn of restored democracy in December 1983. Each service branch owned and ran a domain of large public enterprises focused in the industrial and service sectors.The military had long evaded efforts to separate it from these 45 firms, and was among those who worked closely to block reform under the 1976-1983 dictatorship. Unión Cívica Radical (Radical Civic Union-UCR) President Raúl Alfonsín (1983Alfonsín ( -1989 thought that he could succeed where his predecessors had failed. To his credit, Alfonsín handed most firm presidencies and boards over to civilians. But the military fought off his many efforts to divest it of these same enterprises. As a result, by 1989, the sector was occupied by opposing social forces: military owners and civilian managers. The defense industry was the main battleground in President Alfonsín's struggle to break the power of the military and rebuild democratic institutions. During the Proceso dictatorship (1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983), the military brass split over disputes about leadership, governance, and political strategy. Public condemnation reached a tipping point, meanwhile, over a cascade of human rights abuses and a runaway economic crisis. When military failure galvanized the civilian front and further isolated the military, creating a power vacuum, the democratic forces filled the void 14 La Tercera, "Manejo de sus empresas: El proyecto que incomoda a las Fuerzas Armadas," March 16, 2018. 15 Freer interview. 16 Personal correspondence with the author, July 6, 2018. 17 La Tercera, March 16, 2018. 18 Navarro interview. and commandeered the regime transition. Now, in December 1983, Alfonsín sought to assert civilian over military interests. Rebuking "a profit-seeking and aggressive minority which manipulated the state, regardless of the profound convictions of its personnel," the civilian president intended to seize the military's holdings and put them decisively in civilian hands. 19

Civilian Institutional Strength
The first pillar of Alfonsín's reform effort focused on management. In January 1984, he ordered the Under Secretariat of Production for Defense (UPD) to appoint civilians to the boards of military-owned firms. The military cited the military organic law, which prohibited military from being led by civilians in military institutions, and declared the changes illegal. 20 Yet the bureaucratic context had already begun to change in ways that gave Alfonsín the upper hand. In December 1983, President Alfonsín had obtained a new law from the outgoing military regime (the Ministries Act) creating the UPD, empowering it to name directors and control enterprise reforms, and transferring to UPD's orbit the direction of all military firms. The military resisted, but relented when Alfonsín threatened to go to Congress. A coalition of the main parties had been lobbying the dictatorship for an opening since 1981, making Congress, the brass thought, likely to side with Alfonsín. 21 The dustup set the scene for Alfonsín to exhibit his political strength in bureaucratic corridors, prompting the execution of managerial reform.

Power to Intervene in Sectors
With top enterprise management power vested in its civilian secretary, the civilian bureaucracy proceeded to appoint civilians to the boards of military-owned firms. 22 Facing down military rebellions, it reshuffled to consist of four civilians and two military the board of Dirección General de Fabricaciones Militares (General Directorate of Military Industries-DGFM), an autonomous holding that owned and managed wholly or partly the 13 industries of Fabricaciones Militares and 23 Argentine industrial firms. 23 Then, it reconstituted directories in the mining, naval, forestry, petrochemical, steel, and construction sectors. 24 By June 1984, General [R] Horacio Rivera, who had been president of DGFM, Hipasam, and SOMISA, could lament that "we are now able to understand what happens in Military Industries: By application of [powers granted by] a general law -the new law of ministries -officials without military status have been appointed in the chain of command of the organism, creating a situation that is … < < against nature > > ." 25 As Rivera's remark reveals, management reform succeeded when the civilian bureaucracy became strong enough to impose it administratively on the military.

Military Coalitional Strength
But though down, the military was not out. The second pillar of reform, targeting ownership, appeared in April 1985 via a draft law authorizing the Ministry of Defense (MoD) to create a state-run holding to absorb the military's shares in all 45 firms. The military objected, as did the unions, with the Asociacion de Profesionales Universitarios de Fabricaciones Militares (Association of Professional University Workers of Military Factories-APUFAMI) buying ad space to warn Alfonsín against provoking "adverse reactions in certain sectors." 26 The unions' strength stemmed both from the relatively high level of unionization and their ties to the Partido Justicialista (Justicialist Party, a.k.a. the Peronist Party-PJ), which ruled the Senate. And indeed, the PJ backed labor, demanding that the unions be represented in the firm's directory; the Senate confirm its board, and Congress control its policy. 27 And although the UCR lacked the PJ's tie to labor, it too opposed any policy involving worker layoffs and, thus, also balked. 28 The tables had turned. With the military's political power bound to manifest in the law's defeat, Alfonsín kept his powder dry and abandoned the bill to the gnawing criticism of the mice. 29 The unions and parties the convergence of whose interests benefitted the military disguised their comforts as security and sovereignty concerns. In January 1986, the government vowed to present legislation to privatize the military's biggest steel and petrochemical firms, prompting military cries that these were too vital to growth and security to sell. 30 When one of Argentina's strongest unions, a group of 12,000 defense industrial workers, Unión Obrera Metalúrgica-San Nicolás (Metalworking Workers' Union-San Nicolás-UOM-SM), agreed, the military party coalition in Congress took over and began acting in ways that confirmed the military's political advantage and foreshadowed the bill's defeat. Peronist Deputy Jorge Triaco warned that "we have said in Congress that we totally oppose the measure" (Primo 2006: 178). Meanwhile, the bipartisan Committee of Industry issued two bills effectively eradicating any privatization effort. Chairman Hugo Socchi (UCR) drafted one of them, identifying 29 firms that could only be sold via specific laws and, thus, were beyond the scope of the Executive bill. General Mosconi Petrochemical. 31 The second bill, that of Deputy Jose Manzano (PJ), barred selling firms that "provide a strategic good or service for development or national security." 32 By October 1986, when the Lower Chamber first negotiated Alfonsín's bill, the UCR and the PJ had agreed to exclude the military firms from it. 33 Though the Committee of Industry was not the project's last destination, the PJ-UCR military coalition ultimately blocked it. Thus, in 1985 and in 1986, defense industrial ownership reform failed when the coalition aligned with the military in opposition to reform became strong enough to stop its imposition legislatively by the coalition that favored it.

Argentina, 1989-1997
In Argentina, the first democratic government performed well regarding managerial aspects of defense industrial reform, as exemplified by the appointment of civilians to the boards of military-owned firms. But when Peronist Party President Carlos Menem (1989Menem ( -1999 took power in June 1989, the sector's ownership structures remained untouched and the military did not intend to cede ground. During the campaign, 80% of 110 military officers supported Menem and opposed the UCR's candidate, Eduardo Angeloz, because Menem was not expected to pursue privatization, while Angeloz and the UCR were. 34 Protecting the status quo was a matter of independence and self-sufficiency, said General Isidro Caceres. 35 Menem disagreed and, by 1997, had divested the military of 38 of its firms and turned these over to civilians. Defense industrial reform was a flashpoint in Menem's reform of state and society. In July 1989, Argentine capitalism collapsed: "hyperinflation led the economy to the brink of dissolution, uncontrolled conflicts put the society at the risk of explosion, and lack of governability strained the polity to the limit" (Díaz 2001: 29). Menem took office six months before the installation of the new Congress because Alfonsín had proven unable to manage the crisis and, as Menem had a majority in the Senate but not the Lower Chamber, the UCR Deputies agreed to support any reform bills issued by the executive in Menem's first six months. Embracing privatization, Menem turned to military-owned firms early on because "the public firms that did the most [borrowing] were those involved in investment projects in industries the military favored (especially steel and armaments)" (Frieden 1991: 207). 33 LADB Staff, "Argentine Congress Debates Privatization Legislation," October 14, 1986. 34 Latin American Regional Report: Southern Cone Report, "Argentina's Menem has Little Time to Succeed Where Alfonsín Failed," June 29, 1989. 35 FBIS-LAT, "General Caceres on Restructuring, Other Issues," November 21, 1989. 31 Clarín, "El modelo radical," March 20, 1986. 32 Clarín, n.t., July 20, 1986 Indeed, by 1985, the military and military industry combined had accrued a whopping 3.5 billion-dollar debt. 36

Civilian Coalitional Strength
Menem launched the first salvo in the reform battle in July 1989, sending Congress a bill to privatize, among other enterprises, six of the military's largest petrochemical and one of its steel firms. (This law, the State Reform Law, and with the Economic Emergency Law were Menem's earliest reform bills, and the former was his first privatization law.) Yet the UCR now reserved its right to dissent or alter legislation, and labor-linked Peronists resumed a hostile stance. Menem thus ignored the UCR and allied with small right-wing party Unión de Centro Democrático (Union of Democratic Center-UCD), putting its leaders in his cabinet; sponsored employee share ownership programs to sway labor-linked Peronists; and used exchange relationships to win over small provincial parties. In time, Menem absorbed "a substantial part of the civilian coalition that, in the past, often allied with the military against civilian governments" (Armijo 1994: 29). The PJ-led civilian coalition was geared up to impose the initiative.
The civilian coalition enacted the State Reform Law within a month. Some analysts attribute Congress's approval of this law to the UCR Deputies having agreed to back any reforms issued by the executive in its first six months (Treisman 2003). In fact, the UCR Deputies voted against the State Reform Law and even tried to prevent its defense reforms. On August 11, 1989, a week before the bill's passage, they altered the bill's text to exclude the military-owned firms it targeted. 37 The Peronist Senate reinserted these firms in the text and Menem expanded his support among other parties, 38 but the Radicals then voted against the law (Díaz 2001: 246). Solingen (1998) finds that reform stemmed from the collapse of a statist coalition which had enabled military involvement in strategic industries. This narrative reveals, similarly, that coalition building made the early reforms possible despite, not because of, the UCR Deputies.
The plot thickened on July 23,1990, when Menem sent the Senate the first privatization bill issued since the State Reform Law: one targeting 27 military-owned firms and designating MoD as the application authority. Four days later, Menem dispatched the second: Law 23,809, Privatization of Altos Hornos Zapla (AHZ). The Radicals and one dozen Peronist Deputies, led by the dissident Grupo de los Ocho, which had emerged on the PJ's left flank in December 1989 to contest the privatization program, claimed Congress's right to control the Executive's actions and demanded that the AHZ bidding process be approved by Congress (Llanos 2002: 114-5). Yet the changes in the political context created by the installation of the new Congress reinforced Menem's advantage, as the PJ kept its Senate majority and the PJ, UCD, and provincial party bloc now formed a majority in the Lower Chamber. 36 FBIS-LAT, "Alfonsin, Others Discuss Military Budget," August 15, 1985. 37 Clarín, "Diputados introdujo cambios al proyecto de reforma del Estado," August 12, 1989. 38 El Cronista, "La ley Dromi volvió otra vez la Cámara Baja," August 16, 1989. And so, with little fanfare, the civilian coalition imposed the AHZ law three weeks later (Llanos 2002: 108-9, 126).
The defense industry privatization law proved more contentious. At stake was not just control of military divestiture through public biddings, direct awards, and debt capitalization, but also the power to earmark in later rounds of military reform the billions in expected proceeds from privatization. MoD and its PJ allies (the Rojo-Punzo) claimed control and supported the bill, while Public Works Minister Roberto Dromi and his allies (the Celestes), seeking to retain control they had henceforth exerted over privatizations of non-military firms, allied with the Radicals in Congress and began encouraging expressions of military nationalism. 39 All told, the bill slumbered for 17 months while distributive skirmishes paralyzed the coalition party. Only when Menem made the bill's passage a top priority, in 1991, did the Celestes get back in line. In any case, why would Peronists who opposed privatization under Alfonsín now support it?
The Peronists' reversal followed from the firm control Menem exercised over the party. Menem actively used Peronism's party apparatus to enforce party discipline. By late 1990, he had assumed the party presidency and appointed loyalists to top positions in the party. With control of the party's main patronage resource in Argentina's closed list proportional representation electoral system, its list of candidates for office, Menem proceeded to marginalize opponents of his defense industrial reforms (Teichman 2001: 78).
That capacity to enforce unity within the main party to a multi-party coalition became critical in December 1991, when the PJ-led civilian party coalition within Congress leveraged its power to approve the privatization law. 40 As Armijo finds in studying the timing of privatizations in Argentina and as this narrative reveals, Menem legislated aggressively "even against the interests of the Argentine Army" because of coalitional changes favoring him between mid-1989 and mid-1993 (1994: 27). Thus, in 1989Thus, in , 1990Thus, in , and 1991, defense industrial ownership reform succeeded when the coalition that favored it became strong enough to impose it legislatively on the coalition aligned with the military in opposition to it.

Civilian Institutional Strength
The civilian bureaucrats in the defense ministry confronted military, labor, and private-sector stakeholders opposed to privatization going forward in their companies. Yet the executive bureaucracy had not changed since the days of the Radical government. Like Alfonsín, Menem enjoyed the (now upgraded) Secretariat of Production for Defense's authority to intervene in and control the reform of individual firms. But where Alfonsín turned to party stalwarts to staff the bureaucracy, Menem tapped technocrats, transferring, in February 1991, the Ministry of Economy's team and structure to the MoD. Menem also set aside a World Bank Public Enterprise Reform 40 LAWP, "Military firms cleared for sale," December 19, 1991. 39 Daniel Lupa, "La política militar de la administración," La Prensa. December 2, 1990. Adjustment Loan (PERAL II), a complementary US$200 million loan from the Export-Import Bank of Japan (JEXIM), company resources, and commercial loans that included a Banco de la Nación Argentina loan (World Bank 1996: iii).

Control of Credit and Power to Intervene in Sectors
The civilian bureaucracy used its monopoly of expertise and information, access to credit, and control over enterprise reforms to shape the interests of groups traditionally invested in the military-centered economy. First, it administered compensations that helped to lower the cost of reform for key actors. After Congress had included provisions requiring both, compensations targeted to the military included award of the proceeds from privatization to the military and the inclusion in every tender of clauses for the protection of defense manufacturing (Llanos 2002: 142). Compensations for the private sector included market-share compensation via preferential bidding procedures for partners and suppliers of military companies. Thus, the civilian defense team granted the Army's private partner in Carboquímica Argentina, Ragor, the right to acquire the shares of the firm before bidding opened to other private investors (World Bank 1996: 16).
Compensations targeted to unions and workers formed part of a strategy of labor cooptation, and included market-share compensation via the reservation and direct award of ownership to workers or unions of a military-owned firm undergoing privatization (De Kessler 1993: 138). Menem also allowed the union leaders to administer the workers' shares, which meant that they would earn some of its profits (Etchemendy 2001: 16-18). Subsidies were used generously: to persuade UOM to accept reform, the government forgave $25 million in debts owed by the union's health plan (Madrid 2003: 73). Credit and firm resources, in particular, were harnessed to give workers severance payments more generous than their wages. At SOMISA, where a force of 12,000 would be halved, the average length of the payment was a two year's salary, and the average amount per worker was $15,000 (Kikeri 1998: 24). Meanwhile, the bureaucracy named interventores, trustees who had full executive powers to operate the biggest firms (e.g., SOMISA and PBB) until privatization was complete. Interventores expedited reform during periods of labor militancy as well as became coalition builders in their own right. 41 Second, the civilian bureaucracy amended some earlier proposals to craft changes that would lower the cost of reform by providing an option of early reversal. At shipyard AFNE, it made a tacit deal: in exchange for labor demobilizing, the agency would refrain from selling the firm and transfer it, instead, to the Province of Buenos Aires (Frassa 2010). At Petroquímica General Mosconi, the bureaucracy allayed the union's worst fears by arranging for the firm to be absorbed by the partly stateowned Yacimientos Petrolíferos Fiscales (Fiscal Oil fields-YPF). YPF remained under state management, and the state held veto power over its governance via possession of a golden share (Yi 2008: 135). Negotiations over the fate of aviation firm Área Material Córdoba Fabrica Militar de Aviones did not end in privatization, but with the concession of the maintenance of Air Force aircraft to Lockheed Aircraft Corporation for five years (Da Ponte 2011: 50). That promised to generate large future returns for the state by enhancing the firm's performance before its return to state control, which occurred in 2009.
Between March 1992 and August 1997, Argentina's civilian bureaucracy radically transformed the country's defense sector, seizing and then transferring every big military-owned firm to new private or public civilian owners and managers. The opposition to defense industrial reform of vested military, labor, and private sector interests was no match for an institution empowered to control credit and intervene in sectors. "However, the single most important factor," wrote the World Bank in its completion report for PERAL II, "was the strong personal commitment of President Menem, without whom these reforms might not have taken place " (1996: v). Thus, the policy goal of privatizing military-owned firms succeeded when the civilian bureaucracy emerged strong enough to impose it on the military.
The literature associates military reform with a reduction in the role of the military. In this respect, at least some of the policies described in the Argentine case are unique, including award of the proceeds from privatization to the military and the insertion in every tender of clauses to protect manufacturing for defense. These changes replaced military prerogatives not with civilian, but with other military prerogatives. If civilian prerogatives, such as earmarking of the proceeds from privatization for social spending and the insertion in every tender of clauses to protect civilian manufacturing, do not, in principle, contradict the reform goal, then some of the reforms described in the Argentine case amount to rent allocations. Such changes result in less civilian, that is, in a demilitarization process that generates uneven results across sectors and entails politically obtained market benefits.

External Obligations
It is commonly argued that states obligated to align their civil−military relations with international standards are more apt to pursue military reforms than states not so obligated. Getting leaders motivated to make difficult but not electorally profitable changes is supposed to take being forced to do so by someone else (Bruneau and Trinkunas 2006). So, leaders in states obligated to align defense industry governance with world standards should turn defense firms over to civilians, while states not so obligated should not. Was that so in Chile and Argentina?
In Chile, joining the Organisation for Economic Co−operation and Development (OECD) in May 2010 required agreeing to appoint majorities of civilians to the boards of military−run enterprises in the public sector. The duty had actually been in effect since 2001, when negotiations with OECD began. President Bachelet invoked it in 2007 and then again in 2015 in pushing for her policies. Yet those initiatives, and all such efforts in Chile, failed. In Argentina, no such obligations materialized, but leaders managed to achieve a radical reform.
The evidence does not support the alternative hypothesis. External pressures in Chile should have driven leaders to realize defense industrial reform while the absence of them in Argentina should have kept leaders from doing the same, and yet the opposite occurred.

Strength of the Outgoing Regime
Stronger outgoing regimes can negotiate outcomes more favorable to themselves than those forced out by crisis. Military regimes that ruled more effectively can use their bargaining power to secure a role for officers in the policy process (Agüero 1995). Accordingly, militaries should be better equipped to contest defense industrial reform where the military regimes were stronger than where they were weaker.
In Chile, the military left office strong, confident, and united. It engineered a transition aimed at securing tutelary powers and curbing the powers of civilian governments. As a result, between 1990 and 2005, civilians did not dare to touch the military's firms. Parties of the Left and Right cooperated to strip the military of its tutelary powers in 2005. But despite having lost its tutelary powers, the military managed to fight off defense industrial reform much later. In Argentina, the military lost all powers to control the transition after economic and military failure. We saw, consequently, that President Alfonsín could use a bureaucracy empowered via a law obtained from the outgoing regime to hand military firms over to civilian directors. But if the strength of the outgoing regime alone determined the outcome, then it should have rendered the military powerless to defend its management and ownership of firms. It did not.
The evidence does not fully support the alternative hypothesis. The strength of the outgoing regime did play some role but cannot fully explain the failure of defense industrial reform in Chile. Similarly, the weakness of the outgoing regime played some role but cannot fully explain the partial failure of defense industrial reform in Argentina.

Context in Different Time Periods
Still another standard account is that neoliberal reform under democratic rule creates a context more favorable to demilitarization than one under military rule (Schmitter 1994: 71). Unless the military controls the state, neoliberal reform offers demonstrated successes upon which to create constituencies for advancing defense reforms.
In Argentina, neoliberal reform occurred under democracy and the defense industry was reformed then. And the defense industry privatization law, albeit the first sector−specific privatization bill, passed after a distributive conflict within the civilian party coalition was resolved. Still, it was reforms to the biggest military−owned firms (via the State Reform Law) that launched the privatization drive. The passage of this early law could not have been shaped by the effects of earlier changes because there were no earlier changes. In Chile, neoliberal reform occurred under military rule and the defense sector was spared then and then again under democracy (Lüders 1991: 8). Reforms like those attempted in defense had, however, succeeded in other sectors, as in the case of copper giant CODELDO, but this had no multiplier effect.
The evidence offers only partial support for the alternative hypothesis. The concurrence of neoliberal economics and democratic politics played some role but cannot fully explain the success of defense industrial reform in Argentina. Likewise, the absence of neoliberal reform under democratic auspices in Chile cannot fully explain the failure of defense reform in Chile.

Conclusion
The fate of defense industrial reform efforts cannot be understood without considering the outcomes of coalition and institution building power struggles between civilians and the military. The comparative analysis reveals the importance of differences in relative coalitional and institutional strength. These contrasted with the similarities between the cases of Chile, Argentina (1983Argentina ( -1989, and Argentina (1989Argentina ( -1997. In all three cases, the military elite and certain civilian leaders fought to perpetuate and reform the ownership and management structures of the defense industrial firms built by the military a half−century earlier. They did so in political systems in which a few strong and relatively institutionalized parties fought for seats and well institutionalized and able executive bureaucracies obeyed commands. But these militaries and civilians differed in their abilities to defend or reform the defense industrial firms.
In Chile, the military, unions, and leftist parties' shared interest in preserving the sources of their power allowed the military to overcome its rivals in the races to find supporters and amass decision−making power. Those rivals were forced to watch the congress, bureaucracy, and president and defense portfolio, at different times, defend the military and its control of firms. In Argentina (1983Argentina ( -1989, where labor, partisan, and military interests also converged, neither President Alfonsín's institution nor the military's coalition could dominate the other. Though the former could appoint civilians to the boards of military−owned firms, the coalition could block the institution's bills to divest the military of these same holdings. In Argentina (1989Argentina ( -1997, where President Menem used tactics, finance, and expertise to prevail over the military, reforms passed, civilians intervened, and firms were handed to civilian owners and managers. This paper does not intend to deny some basic tenets of democratization theory or the political economy of reform. It is true that, as a regime legacy and bargaining model would predict, stronger outgoing militaries can last longer in the struggle to ward off defense industrial reform. And it is plausible to say, as do proponents of dynamic approaches, that economic crises weaken actors vested in the military− centered political economy. Overall, moreover, reforming other sectors first (or not) does impact momentum for reform, as gradualist perspectives suspect.
The approach here attempts to capture, instead, the dynamics of civil−military relations in the midst of the defense industrial reform process, stressing the relative strength of the actors. In this more complex picture, coalition and institution formation are individually necessary and jointly sufficient for outcomes. Research beyond these countries and periods is however warranted, and findings in the literature are a good place to start. Studying the Turkish case, Kurç finds that the private defense firms opposed a proposal to unite four military−owned firms under a civilian holding because the new holding was expected to dominate the market. Later, when the government used for the first time its power to appoint civilians to the boards of the military−owned firms, the military could not stop it (2017: 269). Studying the Portuguese case, Barros finds that elected leaders' use of appointments to the boards of defense industrial firms to constitute their political coalitions prevented change (Barros 2005: 96). It would be fruitful to analyze in cases like these whether ownership and management of defense firms was preserved or annulled through coalitional and institutional politics. A wider, cross−regional study could generate stronger conclusions, with implications for development, democracy, and security.
The implications are evident in the Chilean and Argentine cases. First, reform impacted the military's ability to shape large productive sectors. In Argentina, reform effectively freed the steel sector by ending the SOMISA−Propulsora Siderúrgica conflict, which saw SOMISA sell pig iron at below cost to smaller mills in order to keep them competitive and then cover those losses by drawing on the Treasury and overcharging private firms, including Propulsora Siderúrgica, which it nearly bankrupted (Lewis 1992: 267-8). As a result, the steel sector reported large increases in productivity, capacity utilization, and output between 1990 and 1993 (World Bank 1996: 7). In Chile, by contrast, reform failure reinforced the military's monopoly power. When the last private defense firm, Cardoen, left the business in the 1990s, it cited military dominance of the market. 42 It is significant that Chilean advocates of defense industrial reform present such reform as a way to facilitate private capital formation in the defense sector.
Reform also impacted civil−military relations, which affected the outlook for creating domestic wealth via research and development. In Argentina, reform ended decades of quarrels about who controlled what. That made it possible to improve information flow and the quality of human resources. Thus, the Agency of Scientific Research and Technological Development, created in 2007, could administer and fund around 100 research projects between 2008 and 2011 (Government of Argentina 2010: 275). In Chile, by contrast, reform failure, in postponing key decisions about the composition of defense industrial governance structures, must have also delayed choices about creating new institutions to advance the sector. As in Turkey, where, as Kurç (2017) shows, civil−military rivalry prevented the country from reaching its desired goal of defense autarky, the result of these institutional deficiencies was a less dynamic defense industry.
Finally, defense industrial reform or failure impacted access to the global military complex. Reform made it possible for Argentina to benefit from the transfer of technology, technical data, and industrial know−how via integrative relationships like that of Área Material Córdoba's concession of the maintenance of Air Force aircraft to Lockheed Aircraft Corporation. Reform failure prevented Chile from reaping those same benefits via ENAER's planned partnership with European firm Airbus Military and partnerships that defense officials were considering for other firms. 43 The Comptroller's ruling erased any certainty that ventures with multinational defense companies could be launched via administrative means. More comparative studies of defense industrial reform could shed greater light on these implications.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.