Abstract
Comparative constitutional studies rank the US president as relatively weak and most Latin American presidents as strong. However, specialized studies suggest that US presidents have great abilities to implement their agendas. We argue that presidents with weak formal powers “reinforce” their ability to impose an agenda (scope), as well as their ability to make those decisions stick (force). These reinforced powers, however, have diminishing returns as formal powers rise. As a result, the sum of presidential powers ranges from high (the US) to very high (Latin America).
Introduction
The comparative literature brands the US presidency as relatively weak. Shugart and Carey (1992) and Payne, Zovatto, and Mateo Díaz (2007) find that US presidents enjoy fewer powers than their counterparts in a wide range of presidential systems. Similarly, Cox and Morgenstern (2001) place the US presidency and European prime ministers at opposite ends of a continuum that distinguishes the degree to which the executives can integrate themselves into the legislative process. They argue that “Latin American executives typically have greater powers of unilateral action than either U.S. presidents or European prime ministers, but they occupy an intermediate position as regards executive penetration of the legislative process within the assembly” (Cox and Morgenstern 2001: 175).
US presidents, however, are not hamstrung. President Bush was known for his assertion of executive authority, and President Obama has issued executive orders and additional directives to overcome legislative opposition and inaction on issues ranging from economic growth to environmental protection. US presidential scholars recognize these faculties, titling their works The Imperial Presidency (Schlesinger 1974), By Order of the President (Cooper 2002), and Power without Persuasion (Howell 2003). In his review of the US president's unilateral powers, Howell states that “a defining feature of presidential power during the modern era […] is a propensity, and a capacity, to go it alone” (Howell 2005). Krause and Cohen (1997) also show that the president is poorly constrained by the Constitution, revealing that the use of executive orders increases as presidential support in Congress decreases. Recently, there have been concerns that US presidents have attempted to overstep their constitutional boundaries by issuing signing statements in an attempt to establish their intentions and implementation strategies (see, for example, American Bar Association 2006). Is the US presidency significantly weaker than its counterparts in Latin America, or have the comparativists substantially underestimated the power of the US president?
We argue here that the latter is more the case; like coffee sizes that range only from “tall” to “venti,” there are strong or extra-strong, but no weak presidencies. 1 The comparativists, however, are not all wrong in their modeling. Their comparisons of formal powers do suggest important distinctions among presidential systems. Institutionalists have theoretically and empirically shown that differences in veto, decree, and budgetary powers, for example, have important consequences for the executive-legislative power balance (Baldez and Carey 1999; Shugart and Carey 1992; Mustapic 2002; Garretón 1989).
The authors would like to thank Aníbal Pérez-Liñán and the anonymous reviewers for their insightful criticism and comments.
The clash between the conclusions of the American and comparative literature is the result of the case study approach of the former and the limits of larger-n studies in the latter. While presidential popularity and legislative support are available to both sets of studies (Krause and Cohen 1997; Pereira, Power, and Rennó 2005), the independent variables explaining presidential powers generally diverge. The American literature is divided between president-centered (idiopathic) factors – such as a president's leadership style, character, and ability to move public opinion – and a presidency-centered approach that emphasizes institutional capacity and control of the bureaucracy (Hager and Sullivan 1994). The comparative literature, by contrast, uses variance in formal constitutional powers as its primary independent variable. If both sets of experts are correct in their assessments of the degree of power and the variables that explain it, what are the implications for comparative theory about the weak constitutional US presidency and strong Latin American presidencies?
The answer has two parts. First, even if a theoretical model could include weak presidents, empirically the US and Latin American presidents should range from moderate to strong – thus suggesting differences of degree, not of kind. To explore this limited range of the dependent variable, we compare US unilateral tools and Latin American decrees to show the abilities of each in achieving their agendas.
Second, the finding of similar exercise of powers implies that the comparative scales of presidential powers are lacking. If the ability to implement policies is similar but formal/constitutional powers are different, then the scales must be missing important tools available to presidents. At the same time, it may also mean that formal and informal powers are substitutive or complementary rather than additive. In other words, if presidents who are weak on the formal scales have similar abilities to implement their agendas as those endowed with stronger formal powers, it is the former that require greater recourse to informal mechanisms. In short, there are diminishing returns from informal powers when a president is already strong.
We explain the ability of presidents to implement their agendas, regardless of their formal powers, by first dividing the totality of presidential powers into scope and force. These two dimensions of power are shaped by the constitution and other formal rules, which are supposed to define the balance of powers. We then discuss how each of these pieces of presidential powers is “reinforced” with informal powers which are not stipulated and often unwritten and associated powers which emerge from laws or institutions. 2 Examples of the former include the ability to guide public opinion for or against a legislative proposal while examples of the latter would include organizational advantages or the ability to effectively control the bureaucracy. In short, reinforced powers are tools that the executive can use that go beyond constitutionally enshrined mechanisms to directly propose or veto legislation. Oftentimes, the delineations between formal and reinforced powers bleed into each other, complicating strict classifications. US executive orders and some Latin American decree powers, for example, have a constitutional basis, but the limits of authority are ambiguous. Our goal is not to draw strict boundaries between these types of powers, but to suggest that all presidents have recourse to a mix of powers, thus giving them important advantages in the lawmaking process.
We use the term “reinforced powers” to distinguish our concept from literature that discusses informal presidential powers or informal institutions. Our concept is meant to incorporate ideas from these studies, but also include associated powers – indirect constitutional mechanisms that reinforce direct formal legislating or agenda-setting powers.
We begin by summarizing the conventional wisdom regarding differences in presidential power between the United States and the modal Latin American country, focusing largely on their ability to implement policy proposals. The second section introduces the concepts of scope (presidents’ latitude to define legislation) and force (the executive authority to prevent change or abrogation of the executive's policy proposals by the legislature). Section three examines sources of presidential scope, specifically through reinforced powers, and then compares types and volume of unilateral directives between the United States and some Latin American countries. Section four then compares presidential force, focusing on the roles of the legislature and the judiciary. Each of these sections provides clear examples of how formal and reinforced presidential powers create presidents who all have medium or high abilities to implement their agendas. The fifth section concludes.
Powerful Presidencies, Varying Formal Powers
Our motivating question is how presidents possessing different levels of formal powers can all be strong or very strong. The US president's ability to implement policy priorities, sometimes in spite of institutional obstacles, is the focus of many studies. Schlesinger (1974), West and Cooper (1989), and Rudalevige (2005) all argue that presidential intrusion onto legislative turf threatens the theoretical foundations of representative government. Owing to the covert exercises during the Vietnam War and then Watergate, Schlesinger labeled the American president “imperial,” refrains of which were also heard during the George W. Bush administration (Cooper 2005). Meanwhile, Cox and Morgenstern (2001) and Cheibub and Limongi (2002) explain that Latin American presidents intrude into traditional legislative domains, though the branches are not as fully integrated as in the ideal-typical parliamentary system. And O'Donnell (1994) admonishes that overly powerful presidents in Latin America yield a cheapened form of democracy, which he labels “delegative.”
This characterization of similarly strong US and Latin American presidencies clashes with the textbook model and most comparative studies that include the United States as a case. While the US founders were concerned with representation and therefore envisioned a balance of powers between the branches, Latin American constitutional framers were more interested in efficiency and therefore endowed their presidents with greater powers (Alemán and Tsebelis 2005). The resulting constitutions, as coded by Shugart and Carey (1992) and updated by others (e.g. Payne et al. 2002; Payne, Zovatto, and Mateo Díaz 2007; Tsebelis and Alemán 2005) suggest that the US presidency is weak, scoring close to the bottom of scales of presidential powers in terms of absolute scores or in comparison with Latin American countries. The low US score is based on the two-thirds vote required for an override of the US president's veto combined with the lack of a partial veto, decree powers, exclusive initiative, or dominance over the budget. 3
Payne, Zovatto, and Mateo Díaz (2007) code for decree powers, budget provisions, veto and partial veto processes, and the power to convoke plebiscites. Shugart and Carey (1992) also measure control over the cabinet. Tsebelis and Alemán code for veto powers and show their own rankings plus those based on three other studies. In their model, the US president garners a score of 2 on a scale where the median country (Venezuela) is coded with a 7.
This formal weakness, however, clashes with studies showing US presidents with significant success in implementing their agendas (e.g. Howell 2002; Rudalevige 2005; Beckmann 2010). So, how do formally weaker (US) presidents manage to achieve similar policy success as formally stronger (Latin American) presidents? What are the sources of these powers? And what prevents a legislature from protecting its turf?
When answering this first question, authors have commonly focused on US presidents’ “power of persuasion” in legislative bargaining as a function of reputation and prestige (Neustadt 1990), and their ability to appeal to the public and generate a mandate (Neustadt 1990; Lowi 1985; Kernell 2006; Rudalevige 2005; Schlesinger 1974; Canes-Wrone 2005). Ingberman and Yao (1991), for example, argue that public support allows presidents to issue credible veto threats, while Canes-Wrone (2005) demonstrates how presidents’ appeals to the mass public can place pressure on Congress and shift policy in the direction of majority opinion. Other scholarship emphasizes how presidents fortify their bargaining power through control over the bureaucracy and its extensive resources (Neustadt 1954; Cooper and West 1988; West and Cooper 1989; Bawn 1995) and manipulation of both negative (Cameron 2000) and positive (Beckmann 2010) agenda-setting power.
Another branch of scholarship examines US presidents’ ability to unilaterally implement policy without the consent of Congress through such mechanisms as signing statements, national security directives, and executive orders (Howell 2003, 2005; Moe and Howell 1999; Cooper 2005; Mayer 2001; Krause and Cohen 2000). Within the comparative literature, there have been efforts to distinguish different types of presidential powers, such as Haggard and McCubbins’ (2001) classification of proactive and reactive powers. In their view, proactive powers (such as the authority to take unilateral action through orders and decrees) differ from reactive powers (such as the ability to veto or alter legislation through line-item vetoes) with respect to their potential effects on enhancing policy-making by the president.
Informal and associated powers, which we term “reinforced” powers, are less central to studies of Latin America – in part because most of that region's presidents have stronger formal powers over legislative agenda control, including partial vetoes, decree powers, budgetary discretion, urgency provisions that allow the president to force quick legislative action, and the ability to directly initiate legislation (Shugart and Carey 1992; Mainwaring and Shugart 1997; Carey and Shugart 1998; Baldez and Carey 1999; Cox and Morgenstern 2001; Mustapic 2002; Negretto 2004; Tsebelis and Alemán 2005; Payne, Zovatto, and Mateo Díaz 2007). Studies, do, however, suggest that these presidents often have “partisan powers” (Mainwaring and Shugart 1997; Cox and Morgenstern 2001) that allow them to buy or negotiate for the support of legislators (through either legitimate or illegitimate means). However, with notable exceptions such as Calvo (2007), the comparative literature has tended to ignore presidents’ power to persuade the public and their respective congresses.
To explain the interactive role of formal and reinforced powers of the US and Latin American presidencies, we focus on the scope of their power – the range of ideas that they put on the table and can turn into policy – and their ability to ensure that the legislature or courts do not overturn those policies, what we term the force of their power. 4 A president with unrestricted decree authority, for example, would have the highest level of scope. Force, by contrast, is the ability of the president to make those policies stick. For example, if a president issues a decree or executive order, Congress could attempt to change or repeal the requirement with a new law. But because a new law would require the president's signature, the president can usually assure that the decree sticks. Scope would be limited if there were legal or other restraints on the use of decrees, and force would be limited if the legislature could review and rescind decrees.
Our separation of scope and force is similar to previous distinctions in the literature between proactive and reactive powers, but the formal and reinforced aspects suggest an aggregate relationship between the two that increases presidential power overall, rather than positing a trade-off.
Reinforced powers help presidents “fill in the cracks” of formal powers. For different countries, formal scope would include constitutionally listed areas of exclusive executive initiative, decrees, and partial veto provisions that allow the president to implement some but not all parts of bills. These powers over the scope of legislation are reinforced by presidents’ abilities to push legislation through control over their parties, perhaps with the aid of public opinion or the control of budgetary resources that they can direct to recalcitrant legislators. Another factor that may reinforce the scope of a president's powers is the loose interpretation of constitutional provisions – sometimes reinforced by the judiciary which the president frequently controls – regulating decrees or executive orders, signing statements, presidential proclamations and memoranda, and national security directives.
A president's ability to define a policy agenda can also be reinforced by contextual variables and collective action problems that limit the legislature's action. Legislatures in Latin America generally have fewer resources than their US counterparts and must thus rely on the executive for technical expertise in formulating legislation. 5 It seems uncontroversial that this limitation on legislative ability to be proactive further supports Latin American presidents’ scope.
This limit itself could be the topic of investigation. Why do legislatures not vote more resources for themselves?
Presidents also have an array of powers to prevent the legislature from negating their initiatives, which we collectively term “force.” The most important constitutional aspect of a president's force is the ability to veto legislative initiatives. As noted, if veto override requirements are high, then presidents only require a minority of legislative support to prevent legislation that would overturn a policy. US legislators tried to counter this problem by inserting “legislative vetoes” into bills, thus allowing them to oversee and abrogate executive actions that were taken under the authority of that legislation. However, the Supreme Court ruled that legislative vetoes were unconstitutional in the 1983 case of Immigration and Naturalization Service (INS) v. Chadha.
Since the 1930s, the US Congress has frequently written legislation that included provisions in order that it could review and rescind agency rules that met disapproval in either one or both legislative chambers. In some cases, legislation even called for committees to review expenditures or other actions (see Fisher (1985) for examples). Under the Chadha case, however, the Supreme Court decided that such rules violated the Constitution since they were essentially legislative, and as such, they must pass through both houses of Congress and be submitted to the president for approval or veto (Cooper 1983; Leigh 1984). 6 The result of this decision is that while it may be possible to delay the implementation of agency rules (Cooper 1983), Congress would have to either challenge the rules in court or write new legislation, which would then be subject to presidential veto. These authors deal with agency rules rather than executive orders, but presumably Congress lacks a veto over orders and would not have the same dilatory power over orders.
Cooper argues that the logic is flawed, since by making rules, the executive has also failed to follow the same constitutional provisions.
The Chadha ruling is surprisingly similar to Argentina's infamous Peralta decision of 1990, which was brought to the Supreme Court by members of Argentina's Congress who disapproved of President Menem's continual use of necessity and urgency decrees (DNUs) (see Ferreira and Goretti 1998). The Supreme Court – which President Menem had successfully packed – both justified the broad use of decrees (scope powers), and in a manner similar to the US Supreme Court ruling under the Chadha case, reinforced the president's ability to assure that these decrees stuck by arguing that Argentina's Congress could only overturn executive actions with a law. Its decision, however, did not consider the requirement of a presidential signature for any such law.
These examples shows the gray areas of formal and reinforced force, and also highlights the interaction between them. They also show how contextual variables can reinforce both scope and force. Collective action problems, for example, not only give presidents advantages in initiating legislation, they also limit congressional response to executive initiatives. Relatedly, legislators who might wish to overturn a president's action face many veto gates (e.g., recalcitrant committees), especially on issues where the president has legislative allies (Kiewiet and McCubbins 1991; Moe 1993; Krause and Cohen 2000; Howell 2003). Urgency provisions that require short time frames for legislative action may also improve a president's force. In Ecuador, for instance, if the legislature fails to act within a fixed (and short) time on the president's budget or other bills that the president deems urgent, the president's proposal becomes law. Most Latin American legislatures are further hampered by executive control of (or at least sway over) the judiciary (Helmke 2002; Hilbink 2007) and limited access to professional analytical staff that can help them monitor executive actions and write bills (Morgenstern 2006). Short legislative careers and certain types of electoral laws also work against the motivations of legislatures to counteract executive actions (Morgenstern and Nacif 2002).
Table 1 summarizes some of the formal and reinforced powers for both scope and force that we have discussed, separating each by region. The list is meant to highlight tools available to different presidents, but not be exhaustive. For this reason, the table ignores the important distinctions within Latin America, and many of the reinforced powers listed as pertaining to the US presidents are also available to Latin American presidents. Further, as noted, some of the lines are blurred. For example, delegated powers can bleed from formal to informal, as could judicial interpretations of a constitution. The purpose of the table, in short, is to highlight a wide range of powers available to presidents and suggest that US presidents may need to rely more on reinforced scope and force, while most Latin American presidents have both formal and reinforced powers.
Examples of Formal and Reinforced Scope and Force *
Note:
The examples for Latin America are not universal. Among others, decree, urgency, budgetary provisions, and veto powers vary widely, as do the presidents’ control over legislative majorities and the independence of constitutional courts to enforce judicial review.
Source: Authors’ own compilation.
Substitutability of Formal and Reinforced Powers
Because many Latin American presidents have both formal and reinforced powers, it would be tempting to conclude that they are much stronger than their US counterparts. Clearly those presidents who can decree bills with the force of law (Negretto 2004; Carey and Shugart 1998; Pereira, Power, and Rennó 2005), excise individual lines from bills (Alemán and Schwartz 2006), cast amendatory or partial vetoes (Alemán and Tsebelis 2005; Tsebelis and Alemán 2005; Alemán and Schwartz 2006), or demand legislative action through urgency provisions (Cox and Morgenstern 2001) can go beyond what US presidents can do. The question is whether these differences result in separation of presidential power by kind or degrees.
We argue that it is the latter. While there are more excesses in Latin America, studies that suggest that the US president has usurped legislative authority or brand the US president “imperial” support the notion that this weak formal president is not as ineffective as comparative rankings would suggest. Overall, while Latin American presidents enjoy unbalanced constitutions that aid their abilities to direct policy, US and Latin American presidents all enjoy powers that reinforce their ability to set the agenda and ensure implementation of their policies. In this sense, reinforced and associated powers have decreasing marginal utility. When formal powers are at their nadir, reinforced powers prove important in setting the legislative agenda and ensuring implementation of legislation, while extensive formal powers limit the necessity for and utility of associated powers.
Together these analyses imply that while institutions can create a strong presidency, presidents operating in the absence of those institutional prerogatives have other tools to support them. Perhaps then, these two types of powers are substitutes, and/or there is a declining marginal utility of informal powers. In other words, reinforced power is more valuable to formally weaker US presidents than to already powerful Latin American presidents. We support this finding below through an analysis of unilateral actions that shows broadly similar patterns.
Modeling the Scope and Force of Presidential Powers
Figure 1 represents the premises of our model. The first node represents three ideal-typical constitutions that define formal scope powers of a president, covering the range from those that restrict the president's ability to implement an agenda (on the left) to those that provide the executive with strong agenda setting powers (on the right). We depict the US Constitution in the top figure, using a bold line to suggest a relative balance of power. We depict the ideal-typical Latin American president in the bottom figure, using a bold line sloping to the right. The next set of nodes adds the powers that reinforce scope. Since reinforced powers are typically positive, we have used bold lines to imply that presidents with weak or neutral formal powers will likely range from moderate to very strong after adding these reinforced powers. However, the bold line implies only maintenance of power for the powerful presidents in order to suggest that where presidents already have strong formal powers, reinforced powers are superfluous. In short, the graph is meant to highlight the idea that formal and reinforced powers are not simply additive; instead, reinforced powers are – at least in part – substitutes providing decreasing marginal utility as a president's formal powers increase.
The Scope and Force of Executive Powers
We model a similar dynamic with reference to force. Again, formal powers can be positive, neutral or negative. We use a bold downward-pointing line in the top figure to depict a president under a constitution that supposes a balance of power, while the bottom figure has a bold line sloping to the right to represent a president with stronger constitutional powers. The justification for the differently sloping lines is the same as for scope.
These two different paths lead to a similar outcome: powers within a small range on the right side of the graph depicted by the gray oval. Presidents with weaker formal or reinforced powers – in relation to scope and/or force – might end up on the left part of the oval; those with stronger power, to the right side. Overall, however, presidents will be tall and venti, but not small. To empirically ground this finding, the following sections examine reinforced scope and force in US and Latin American cases to show how executives use formal and reinforced authority to implement their agendas.
Putting Policy on the Table: Formal and Reinforced Scope
Here we summarize types and give examples of unilateral executive actions, beginning with formal decree powers in Latin America and then moving to the reinforced powers often employed by US presidents. Using these types, we then compare the frequency of unilateral directives across the United States and a sample of Latin American countries. Using the volume of directives as a proxy, we show that Latin American presidents are indeed extremely strong. While formally weak US presidents do not reach the levels of agenda-setting power enjoyed by their equivalents in Latin America, they do retain considerable unilateral powers.
Latin America
In Latin America, the combination of constitutional provisions and control over political parties has generally translated into high passage rates in the legislature for president-initiated bills – although at times it has also fallen as low as 10 percent in countries as diverse as Brazil, Costa Rica, and Ecuador (Alemán and Navia 2009; Saiegh 2009, 2010a). Here we focus on a different aspect of scope: these presidents’ abilities to skirt the legislative process and implement policies unilaterally (Tsebelis and Alemán 2005; Carey and Shugart 1998). Such powers are sometimes enshrined as constitutional decree authority (CDA), and in some cases this is augmented by delegated decree authority (DDA). Furthermore, informal mechanisms have helped presidents to interpret vague constitutional language that references the decree power in their favor. Pereira, Power, and Rennó (2005) explain, for example, that while the Brazilian Constitution has limited decree power for presidents, it also allows presidents to issue “provisional measures” (medidas provisórias; MPs) for urgent issues. As a result, presidents have, with the tacit tolerance of the Brazilian Supreme Court, taken liberties with the definition of the term “urgent” to implement policies that were not necessarily timesensitive. In his time in office, for example, President Fernando Henrique Cardoso relied on MPs to set regulations on professional sports teams (MP 39) and reorganize the federal bureaucracy (MPs 37 and 38). A second example regards the constitutional limit on the effectiveness of MPs to 30 days, which has seen presidents simply reissue them once they have expired. Although the Brazilian Congress and President Cardoso eventually ended this practice, the example shows how presidents have followed multiple and perhaps questionable routes to implement their policies.
Latin America is replete with similar examples. President Menem of Argentina (1989–99) was among the most notorious of the decree-wielders, in spite of a constitution that is mute with regard to such powers (see Ferreira and Goretti 1998; Mustapic 2002). As noted, a congressional attempt to challenge this practice backfired when the packed Supreme Court reified the decree power. Moreover, Menem's use of decrees suggests he had little respect for the legislative process. For example, he changed the Organic Law of the Central Bank, froze lawsuits against the National Social Security Administration, reduced budgetary outlays, ordered the televised transmission of soccer matches, and set requirements for the payments to social security from soccer revenues.
In response and as part of a deal to permit presidential reelection, the Argentine Congress forced a constitutional change intended to rein in the use of presidential decrees. Specifically, a formal mechanism was brought in to allow the legislature to oversee and regulate decrees. Despite this, presidents continue to use decrees as a standard tool. Since Menem's presidency, Argentine presidents have never issued fewer than one thousand decrees in a single year – and it is often double that number.
A similar dynamic has taken place elsewhere. Democratically elected Venezuelan president Hugo Chávez often relied on constitutional DDA despite enjoying a legislative majority through most of his time in office. For example, in mid-December 2010, Chávez took advantage of his expiring supermajority in the lame-duck National Assembly to obtain passage of an 18-month-long enabling law (ley habilitante), the fourth time he was delegated this power since coming to office in 1999 (Daniel 2010; Corrales 2011). This law, which was denounced by the US Department of State as autocratic, allowed him to issue decrees across a wide range of areas including housing, land, finances, and security. Chávez used decree powers to pass hundreds of laws, including measures to nationalize parts of the oil industry (Decree 5200) and expropriate businesses and private lands (Decree 7915). However, similar to Menem, decrees were also used for unessential business, such as prioritizing the importance of the Internet for national development (Decree 825) and naming citizens to positions in low-level bureaucratic agencies (Decree 8483). As is the case across Latin America, the high volume of executive decrees obfuscates the much smaller number of decrees that clearly aim to legislate.
While Menem and Chávez are infamous for their unilateral governing styles, even presidents perceived to be pluralistic democrats rely on executive decrees in a combination of vital and non-vital policy areas. In particular, Chilean presidents issue dozens of decrees per year and Uruguayan presidents regularly proclaim over 500 decrees per year. Most of these deal with non-vital and non-urgent matters – such as declaring official mourning for the death of Argentina's ex-president Raúl Alfonsín (Uruguay Decree 166/009) – but others are quite significant. One notable substantive decree in Uruguay determined welfare eligibility (Decree 118/005) and an example from Chile was the Supreme Decree authorizing the distribution of the so-called day-after pill to females over 14 years of age without parental consent (although this was later overturned by the Constitutional Court).
Finally, in Panama, the president's scope is reinforced by the constitutional provision allowing cabinet ministers to issue decrees (Decretos Ejecutivos). Because these ministers are themselves appointed directly by the president (through decree), their legislation is bound to be directly or indirectly influenced by the president.
In sum, there are multiple examples of Latin American presidents using formal and reinforced powers to implement policy unilaterally. And though we have not dealt with them here, many studies highlight these presidents’ success in controlling the legislative process. The presidents’ scope, therefore, is very broad. Some presidents are empowered through more generous constitutions, while those without strong constitutional decree powers still issue them frequently. Many of these decrees have wide-ranging impacts; others simply show the presidents’ latitude in implementing policies without legislative input.
The United States
US presidents are formally decree-less, but the scholarship examining US executive orders (see Deering and Maltzman 1999; Howell 2003; Mayer 2001; Krause and Cohen 2000) shows that they too have had success in implementing their agendas. As Howell (2005) explains, US presidents have two fundamental ways of advancing their policy agenda: submitting proposals through Congress or exercising their unilateral powers. These directives – which straddle a line between formal and reinforced scope – include executive orders, executive agreements, proclamations, national security directives, recess appointments, and memoranda; they allow the president to dictate policies or manage the bureaucracy without congressional endorsement. 7 Another tool is central legislative clearance, which although in place since the 1920s was significantly strengthened by President Reagan's Executive Order No. 12291 (West and Cooper 1989). This requires that agencies submit their rules to the president through the Office of Management and Budget (OMB) before publication. 8 The procedures not only allow presidents to oversee bureaucratic procedures, but also to apply political litmus tests to agency decisions (Gilmour 1971; Larocca 2006). In so doing, executive rulemaking crosses into the legislative arena.
Additional types of orders typically associated with administrative rather than legislative matters include presidential determinations and notices.
West and Cooper (1989) also note that the OMB lacks the legal authority to force compliance by agencies with these procedures, and there have been examples where the agencies have not followed OMB requests. The authors, however, argue that the president's influence over the agency heads is generally sufficient to ensure compliance.
In contrast to Latin American countries, there is no constitutional provision or statute that explicitly permits executive orders, apart from rather vague language granting “executive power” (Article II, Section 1). These orders, however, are often of high consequence and can only be overturned by legislation – which is subject to a presidential veto – or court proceedings, implying both high force and scope. Howell shows that, in most cases, presidents use orders to change existing policies over which Congress remains gridlocked (Howell 2005, 2003).
Beginning with George Washington (see Nelson 1989), US presidents have issued orders dealing with key political and social issues. Between approximately 1980 and 2010, presidents issued between 40 and 70 orders per year. However, earlier presidents, especially Franklin Roosevelt and Harry Truman, issued many more (see Table 2).
US Executive Orders
Source: National Archives and Records Administration n.d.
Most of the orders created task forces, set up commissions and advisory boards, or implemented congressionally approved programs (Warber 2006). A large percentage of the orders, however, crossed into the realm of legislation.
Dividing the orders into administrative, symbolic, and policy-oriented categories is an inexact science, since (as Madison noted) the boundaries between the branches are “in general so strongly marked in themselves, [they] consist in many instances of mere shades of difference” (cited in Fisher 1978). It is arguable, therefore, whether bureaucratic reorganizations, pay-rate changes, or even the setting up of investigatory commissions fall squarely into the administrative category. Still, it is necessary to consider the content of the orders and classify them in order to discuss intrusions of the executive onto the legislative turf. West and Cooper (1989) find an impressive growth of “quasi-legislative” rulemaking in the 1960s and 1970s. This is supported by data from Warber (2006), who classifies all US executive orders from 1936 to 2004 as symbolic, routine, or policy-making.
The policy content of these orders is evident. Bill Clinton, for example, issued orders that blocked property and prohibited transactions with the Taliban (Exec. Order No. 13129), declassified documents (Exec. Order No. 12937), affected child support enforcement (Exec. Order No. 12953), and changed federalist arrangements (Exec. Order Nos. 13083 and 13132). Table 3 provides examples of the executive orders issued by George W. Bush and Barack Obama with legislative content. This list of orders includes those that affect the economy, foreign relations, diplomacy, and security. A statement by President Obama clearly indicates that he views this type of power as a tool to overcome legislative obstacles:
US Executive Orders with Clear Legislative Content – Presidents Bush and Obama
Source: National Archives and Records Administration.
In recent weeks, we decided to stop waiting for Congress to fix No Child Left Behind, and decided to give states the flexibility they need to help our children meet higher standards. We took steps on our own to reduce the time it takes for small businesses to get paid when they have a contract with the federal government. And without any help from Congress, we eliminated outdated regulations that will save hospitals and patients billions of dollars (24 October 2011: Remarks by the President on the Economy and Housing).
Other reinforced tools of direct administration are even less clearly defined than executive orders. Presidential memoranda are pronouncements by the chief executive that are similar to executive orders in content and use, but lack guidelines for issuance and publication. Often, memoranda are directed at executive agencies to establish policy guidelines. A similarly vague tool is the presidential proclamation – an instrument that states a condition, declares a law, recognizes an event, or triggers the implementation of a law (Cooper 2002). Some of these are ceremonial (such as the declaration of national days of observation), while others are quite substantive (such as Washington's Proclamation of Neutrality or Lincoln's Emancipation Proclamation).
A last instrument of associated direct presidential action is the signing statement (Garber and Wimmer 1987; Cooper 2005). This pronouncement, which gained prominence under the Reagan administration, can be issued by the president at the moment of signing a congressional bill. Presidents use the opportunity to provide general commentary on the bill – for example, interpreting the law, announcing constitutional limits on its implementation, or indicating how bureaucrats should administer it (Cooper 2005). In practice, signing statements have been employed as a type of line-item veto to set boundaries on the reach of legislation and to structure its implementation, but without the use of the formal veto or the opportunity for legislative override processes. From the president's perspective, the strength of these pronouncements is that there is no constitutional or legal impediment to their issuance, thus disabling the legislature's ability to respond.
In sum, it is clear that US presidents change policy in important ways without congressional action. Figure 2 shows a breakdown of direct presidential action in the United States between 1993 and 2011, adding these four reinforced powers: orders, signing statements, proclamations, and memoranda. The yearly totals surpass 200 and approach 300 in most cases, indicating a much more active US president than commonly assumed.
Unilateral Presidential Directives in the US (1993–2011)
Comparing the Frequency of Unilateral Directives
The above examples are meant to illustrate that while many Latin American presidents have more formal power (and many have arguably abused their reinforced powers), US presidents also have sufficient power to implement important policy items. As a result, US presidents only look weak on a relative scale, not on an absolute one.
The frequency with which different presidents use unilateral powers suggests a similar interpretation. Figure 3 compares the total number of unilateral directives (executive orders, presidential proclamations, memoranda, and signing statements) issued by US presidents with all executive decrees issued in eight Latin American countries between 1993 and 2010. While US presidents issued between 200 and 300 directives per year, the Latin American presidents issued far more: between 400 and 900 for most country-years with the exception of Nicaragua (fewer decrees issued) and Argentina (more decrees issued). While quantity of decrees is not proof of their quality, it seems reasonable to assume that Latin American presidents manifest broader decree authority than US presidents. 9 Still, the very high number issued by the US presidents reveals the broad authority of that office.
Comparison of Unilateral Directives in the US and Eight Latin American Countries
These figures underestimate the Latin American unilateral actions because they do not account for additional reinforced powers available to Latin American presidents, which have not yet been systematically collected. The main point, however, is still clear: while many Latin American presidents use unilateral authority more often, US presidents also issue a large number of directives.
Whether or not US executive orders and Latin American decrees differ in degree or type, these data do highlight the high frequency with which the respective executives direct policy without input from the legislature. As we discuss elsewhere, legislatures can constitutionally review these edicts, but presidential veto power and (frequent) judicial unwillingness to oppose presidential authority can cement these executive initiatives into law. In other words, the wide scope presidents enjoy through their control of agency rules and “administrative” decisions is further enhanced by the constitutions, the courts, and the organizational challenges faced by national legislatures.
Keeping Policies in Place: Formal and Reinforced Force
Regarding force, the modal Latin American president enjoys stronger formal powers than their US counterparts – but the latter have been able to shore up their powers through reinforced mechanisms. To a greater degree than scope, the concept of force lends itself more easily to manifestations of both positive and negative reinforced power; this is illustrated most clearly by potential constraints issued by legislatures and the judiciary. We explain this with a focus on each institution, first in Latin America and then the United States.
Limits on Overturn Attempts by Latin American Legislatures
Many Latin American constitutions support their presidents’ ability to fight off legislative attempts to overturn executive initiatives, while several associated powers provide an extra buffer. In particular, many Latin American presidents have partial veto powers – some can limit the time for legislative action, whereas most can prohibit the legislature from making any significant changes to the budget. The two most important tools on the reinforced side are control of parliamentary parties (which helps to block legislation or veto overrides) and influence over the courts (which have limited authority with regard to executive actions).
The partial veto is a tremendous weapon available to presidents in at least 12 Latin American countries (Saiegh 2010b: 56). 10 It is a much more potent than the blunt package veto, because it allows presidents to undo (and therefore stop before they are started) legislative logrolls. For instance, if the president excises a legislator's favorite project, then that legislator has little incentive to enter into a deal with a second legislator. In sum, the partial veto complicates dealmaking and should thus severely limit the ability of legislatures to solve collective action problems and overturn presidential edicts. 11
Payne, Zovatto, and Mateo Díaz (2007) ascribe the strongest partial veto powers to Argentina, Chile, and Ecuador. Tsebelis and Alemán (2005) indicate that presidents in others countries also have strong partial veto powers. For example, they explain that the Brazilian president can promulgate nonobjectionable parts of a bill. The partial veto can only be overridden by a majority of the House and Senate, who vote jointly.
The partial veto can have a positive side. Samuels (2002) credits the elimination of much pork-barrel spending in Brazil to the partial veto, even though legislators are each empowered to submit amendments worth about USD 1.5 million (see Ames 2002). Our point here, however, is that the partial veto prevents legislative action.
At least in one case, ambiguous language even turned the reactive partial veto into a proactive power. In Uruguay in late 1994, President Lacalle vetoed an article from the new Organic Law of the Central Bank and sent the bill back to the legislature with a “suggested” alternative version of the crucial article. Congress was unable to muster a quorum to debate the veto and, as a result, the published law included the president's substitute article – an article that had never been voted on by the legislature. Constitutions in other countries also define partial veto authority in language that is ambiguous enough to allow Uruguayan-style suggestions to become law. 12
On “amendatory observations” see Alemán and Tsebelis (2005).
A second friend of Latin American presidents is the time limitation that some constitutions impose on legislative action with regards to the budget and bills the president deems urgent. To take just one example, under both the 1998 and 2008 Ecuadorian Constitutions, executive-initiated bills of “economic urgency” can be approved, modified, or rejected by a congressional majority. However, if a bill is not ruled on within 30 days, it officially becomes a decree-law (decreto-ley).
Judicial Upholding of Executive Actions in Latin America
As an additional reinforced power, Latin American presidents have frequently been able to control the judiciary – although some are becoming more independent. 13 The Argentine Supreme Court, which validated military decrees and abstained from political questions, exemplified the old model (Nino 1993; Larkins 1998). 14 Even recently, Helmke explains that Argentine executives routinely cleaned house in the Supreme Court, resulting in justices fearing for their jobs (Helmke 2005). 15 Courts in some countries, however, have recently gained greater independence and power of judicial review, leading to a high degree of variation in the ability of constitutional courts to arbitrate policy decision (Helmke and Ríos-Figueroa 2011). One noteworthy case was the Colombian Constitutional Court's 2010 ruling against a referendum that would have given citizens the ability to allow President Uribe to seek a third term in office. Similarly, the Constitutional Chamber of the Costa Rican Supreme Court has gained a high level of independence over the past twenty years (Wilson 2011; Wilson, Rodríguez Cordero, and Handberg 2004). Simultaneously, in countries including Ecuador and Bolivia, courts with counter-intuitively independent judges have now lost much of that independence as presidents fight back (Basabe Serrano 2012; Castagnola and Pérez-Liñán 2011). Like the recent provisions that have tried to regulate presidential use of decrees, this growing independence of courts in some cases perhaps suggests a growing maturation of democracy and a concern for a more even balance of powers – though the effects are far from uniform across the region. However, presidents’ resistance to the changes indicates that while they may not retain outsized powers, they will still keep enough force to prevent frequent overturns of their policies.
We thank the anonymous reviewer who emphasized how this relationship has recently evolved.
Nino also highlights the limited period of democratic rule in Argentina.
See Larkins (1998) for a review of Menem's Supreme Court packing.
Limits on Overturn Attempts by the US Legislature
US presidents lack the partial veto or the ability to impose time limits on legislative action that help Latin American presidents prevent actions that could counteract their decisions. Why then, in the face of a significant number of unilateral decisions by US presidents, has there been little legislative response to many of these intrusions? We contend that even though the US president lacks some advantages enjoyed by presidents of Latin America, a combination of formal and reinforced provisions limit legislative reaction.
From the perspective of Kiewiet and McCubbins (1988) and McCubbins and Schwartz (1984), the US Congress tries to write legislation in ways that allow agencies broad latitude in areas that serve the legislature's interests or limit freedom in other areas (statutory control). Congress, they argue, also relies on monitoring mechanisms (“police patrols” and “fire alarms”) so that it can react if the bureaucracy acts in ways that harm the congressional majority's interests. Once the legislature discovers prejudicial action, it can threaten budget cuts or more limiting legislation. West and Cooper (1989) add that Congress has significantly increased the resources and staff dedicated to oversight to accomplish these goals.
This perspective has several problems. First, while the problem is surely much more severe in Latin America where legislators still often work with no more than a private secretary, even in the US Congress the lack of resources means that “any such review [of the multitude of rules] inevitably was sporadic and haphazard” (Pierce Jr. 1985: 483). Furthermore, like all organizations, the US Congress faces a collective action problem that limits its ability to legislate in a timely manner.
A formal mechanism, the veto, is another part of the answer. Fisher (1978: 244) puts the problem most clearly: while Congress “delegate[s] authority by a majority vote [it can] recapture it only with a two-thirds majority.” 16 In short, even if the legislature uncovers executive decisions that it wishes to undo, it must rally two-thirds of its members to take action. As important as this limitation is, it has been much more acute since (as noted earlier) the US Supreme Court issued its ruling in INS v. Chadha in 1983, which ended the use of legislative vetoes.
A keen example regards the “gag rule.” In Rust v. Sullivan (1991) the Supreme Court upheld the Department of Health and Human Services’ questionable interpretation of the statute to prevent clinics that receive federal funds from counseling with reference to abortion. Congress wrote new legislation to overturn the rule, but President Bush vetoed the bill (Clayton 1994).
Berry's (2009) interesting work on the Chadha decision illustrates that despite the ruling, Congress has continued to use legislative vetoes extensively. Using a carefully constructed database, he also shows that the president uses reinforced powers – this time in the form of signing statements – to overcome unwelcomed legislative oversight. 17 The president is not always successful, but Berry shows that (a) while the Congress continues to insert legislative vetoes, in part because the court ruled them invalid they often fail to take effect, and (b) bureaucracies sometimes react to presidents’ signing statements rather than the enacted legislation.
Berry argues that presidents use signing statements to counteract legislative vetoes (and there is thus a positive correlation between legislative vetoes and the issuance of signing statements).
Judicial Upholding of Executive Actions in the United States
While the veto and resource restraints provide US presidents with significant advantages, they do not have enough control over the courts to prevent a determined legislature from overturning unilateral actions. Although the US Supreme Court dealt a severe blow to the legislature when it struck down the legislative veto, there are other rulings that limit executive actions – especially since there is a supposition that the Supreme Court will not allow the executive to directly contradict legislation.
An example of this limitation can be seen in the challenge to George W. Bush's issuance of Exec. Order No. 13233, which sought to regulate the Presidential Papers Act. The aforementioned act explicitly states that the president's papers are to be turned over to the National Archives and Records Administration for publication no more than 12 years after the president leaves office. The first affected presidency was to be that of Ronald Reagan, and the papers should have been released in 2001. However, President Bush issued an order stating that the papers would only be published after both the current and affected president gave their approval. This clearly contradicted the letter and spirit of the law, yielded bi-partisan reproach in Congress, and was challenged in court by Public Citizen (a nonprofit organization) on behalf of the American Historical Society. The Supreme Court's ruling favored the opening of the records, and most of the papers have since been published. Nevertheless, the Supreme Court only struck down a portion of the bill, and it took a second order by President Obama in 2009 to revoke the original order. In sum, while this case shows that the president is not without constraints, it also shows the limits on the legislature's reactions.
Conclusion
In determining policy outcomes, constitutions vary in the endowments they grant their presidents. Beyond these formal powers, presidents also enjoy associated and informal tools to help them implement their agendas. Policies are kept in place through mechanisms such as presidents’ partisan powers, a lack of resources in the legislature, and collective action problems that limit legislatures’ abilities to resist presidential initiatives. In this paper, we argue that these informal and associated tools reinforce presidential power in bargaining games. Furthermore, while informal and associated powers are important everywhere, we argue that decreasing marginal returns imply that they take special import where constitutions (like that of the United States) grant presidents only weak formal powers.
Our review of unilateral actions by executives in Latin American and the United States shows that while many presidents in the former do in fact have stronger formal and reinforced powers, the US president also has many tools and advantages over the legislature. If the constitutionally weak US president has such powers, then perhaps comparative studies should recalibrate their scales, focusing only on a range of medium to high.
Since there seems to be a general consensus that concentrating too much power in the presidency is dangerous to the quality (if not the stability) of democracy, several countries have already initiated reforms to address the executive-legislative imbalance. Several areas seem ripe for further reform. First, legislatures require more resources and further professionalization. This perhaps justifies dedicating international aid and domestic efforts to both modernizing legislatures’ technical capabilities and training legislators and their professional staffs. Still, the US experience suggests that even with extensive resources, oversight might be haphazard.
Second, both the United States and Latin American countries should reconsider the issue of legislative vetoes. Even if the legislature has the capacity to oversee executive decisions, this often requires a veto-proof majority in Congress to overturn them – meaning the balance is decidedly in favor of the executive. Since there is also a concern that presidents need the ability to act quickly in response to emergencies, a good starting point might be the Brazilian decree rule – whereby Congress must either convert the decree to law or it loses validity. That type of law could facilitate the desired governability without significantly degrading the representative system.
Third, the budgetary limitations on many Latin American legislatures are incompatible with the separation of powers. If legislatures are prevented from even considering how to collect and distribute public moneys, then the legislature becomes little more than a forum for debate or a rubber stamp.
Fourth, the lack of judicial independence and judicial review has helped Latin American presidents consolidate power, suggesting that judicial reforms could help to limit presidential powers. Many Latin American countries have undertaken significant judicial reforms, and some courts now do have the power to negate laws and decrees. However, without strengthening the legislature and eliminating the restrictions on their abilities to legislate, the judicial reforms cannot redress the executive–legislative balance.