The significance of the distribution of ministerial portfolios for the performance of governments has been extensively analyzed in the literature on parliamentary democracies (Laver and Shepsle 1990; Austen-Smith and Banks 1990). Despite some authors' contentions that research on coalitions in presidential systems is rare (Altman 2000), there is a growing and rich literature developing on portfolio distribution in presidential democracies. Mainwaring and Shugart (1997, 38) explore how cabinet changes are used as "safety valves" in facing crisis, and Stepan and Skach (1994) analyze how presidents can use cabinet appointments preemptively to secure cooperation (1993). Amorim Neto's (2006) path breaking work finds that such coalitions can aid presidents in their policy-making strategies. Similarly many other studies (Lanzaro 2001; Deheza 1997; Altman 2000; Amorim Neto 2000, 2006; Martinez-Gallardo 2005) find that cabinet formulation strategies can help presidents overcome the challenges associated with what had been identified as the uncomfortable and purportedly unworkable combination of presidentialism with multiparty systems (Valenzuela 1994; Mainwaring 1993, among many others).
While some studies are breaking important ground on the significance of the distribution of ministerial portfolios, there is virtually no work outside of the United States on the less formal aspects of executive branch structure. As Bonvecchi and Sartascini's comprehensive review of the literature notes, despite extensive research on coalition politics and economic and political decision making in presidentialism "no equivalent knowledge exists about the organization of the executive branch" (2011, 145). What is more, of the little work that exists, virtually none of it focuses on informal advisory networks or how they can help to make such cabinet sharing schemes function or malfunction in practice. The importance of informal advisory networks, though well analyzed in the literature on the U.S. presidency, is virtually nonexistent in the literature on Latin America. Nonetheless, anecdotal evidence from across countries suggests that informal advisers exist, are often overlooked, and can be quite influential. Indeed, as noted by Link and Glad, "[t]oday the personal advisers to the United States President often have greater impact on American policy than do cabinet heads" (1994, 461). Though little recognized, this is also likely to be the case within the context of Latin American presidentialism. This raises important questions for students of Latin American presidencies: what are the strategies that presidents use to deal with the tensions between formal and informal advisory networks? What impact do they have on policy-making and portfolio-sharing arrangements?
The literature on informal advisory networks in the United States is a logical point of departure for the study of the topic in Latin America, yet the U.S. two-party system provides little insight into how these tensions are managed and magnified in multiparty systems. In terms of coalition management, because cabinet ministers also provide some advisory functions, the way presidents balance portfolio sharing with their need for information and advice in a multiparty coalition is likely to be quite different. In terms of magnification of tension, a whole new set of differing incentives for presidents, ministers and advisers is introduced when they are on separate partisan teams. Fundamentally, multiparty presidentialism introduces an additional element of tension between formal and informal advisers when it comes to presidential control over policy and coalition management. These two goals may also often work at cross purposes in a way that they do not in the United States or under single party government, as this article will show.
This article provides a comparative qualitative study of how presidents deal with the tension between formal and informal advisory networks in multiparty presidential regimes, noting the differences from how these tensions are managed in single-party governments. It does so through a case study of the first four postauthoritarian governments in Chile. It begins with a brief literature review of presidential goals and strategies for achieving them through advisory networks. It goes on to provide a discussion of case choice and methods, followed by a brief theoretical discussion of the tension between formal and informal advisory networks. This section is followed by in depth case studies for the first four postauthoritarian governments in Chile, analyzing both the formal distribution of cabinet portfolios and the less formal aspects of executive branch structure, consistently underscoring how the dynamic of achieving presidential goals was affected by the exigencies of coalition management and maintenance. In particular, as the four post authoritarian governments unfolded, presidents gradually came to rely less on advisers within the formal structures of government (and in particular ministers) and more on informal networks outside the ministries. This resulted in the gradual shifting of responsibility and influence away from ministers toward more informal networks of advisers, which in turn undermined the very authority and influence of formal ministry appointments on which power-sharing arrangements are predicated. The conclusion recounts the study's findings and contributions.
Though fundamentally a case study, this article provides important theoretical lessons for the comparative study of presidential advisory networks. First, the case demonstrates the multiparty coalition management inserts unique tensions into the presidential advisory equation and thus unique challenges for presidents in achieving their governing goals when compared to single-party government. Second, it points to a potential fault line in such cabinet sharing arrangements that can lead to deterioration in their effectiveness: the interplay between the formal distribution of cabinet portfolios and the less formal networks of presidential advisers. This key, but overlooked, variable can help determine the success of portfolio sharing arrangements. The central irony is that the process of sharing cabinet portfolios limits the range of potential presidential appointees, so presidents may be tempted to rely on advisers outside of ministries. On the one hand, these less formal networks of advisers may be intimate aides that help to protect presidents and allow them to promote their agendas. On the other hand the influence of informal networks (1) (often tied more closely to the president and his or her party and located outside the ministries) has a cost in terms of the expertise of ministers and presidential access to information. It also may undermine the very principle of cross-party consultation on which the distribution of cabinet portfolios is based.
Formal and Informal Advisory Networks in Presidentialism
The literature on presidential advisory networks in the United States demonstrates the myriad ways presidents deal with the natural tension that grows out of presidents' goals of gaining the information they need to ensure that policy decisions are made and enforced while simultaneously ensuring the loyalty of their appointees. Lewis' (2008) work points to the tendency of American presidents to politicize the bureaucracy by inserting their own supporters in an effort to ensure loyalty, especially where bureaucrats' career goals may be at odds with the policy goals of the president. However by doing so, presidents may not receive the specialized information they need from bureaucrats who have policy expertise, creating a tradeoff between agency wins and information losses. So presidents may politicize the bureaucracy to increase agency wins, but often at the cost of information losses.
Terry Moe (1985, 1994) has analyzed similar phenomena, but focuses on the creation of presidential agencies. Moe contends that presidents may increase the number of presidential agencies, dubbing an increase in the number "centralization." For Moe, centralization is an effort to shift the functions of the bureaucracy into the core executive. On the other hand, presidents "decentralize" when they move agencies from the core executive and place them back under the authority of a cabinet minister. Rudalevige similarly focuses on another presidential strategy of centralization, which is simply "the shift of duties and functions from the wider executive branch to the Whitehouse staff' (2002, 6). Fundamentally, these decisions are based on how much confidence presidents have that their Secretaries, Ministers, and agency heads will loyally design and implement policies in line with the preferences of the president.
All of this literature is based on the premise that the president will staff her cabinets, the bureaucracy and presidential office with party loyalists, given the single-party nature of U.S. governments. However, multiparty government introduces an additional element of strategic complexity to the traditional duopoly of information and loyalty that are usually underscored as the two basic needs of the president. Coalition management now is added to the equation and becomes as important an imperative for presidential success. For example, Chilean presidents certainly politicized the bureaucracy in the way that David Lewis argues for the United States. However, it is a different type of politicization as a result of multiparty cabinets and the exigencies of coalition maintenance--appointments reflected the overall partisan makeup of presidents' coalitions, not their parties. In this sense, appointments were based on a logic of politicization, but forced politicization to reflect party balance in the coalition. So this different brand of politicization meant presidents could not use appointment powers as a way to appoint party loyalists in the way presidents do...