The number of expert appeals to strengthen state institutions in Latin America matches only the wealth of studies and news reports documenting their failures. Corruption plagues state agencies across the region (see Berríos Reference Berríos2010). The Weberian ethos of devoted public service, where it occurs, is the exception, not the norm. Even in the absence of graft, embezzlement, patrimonialism, and other illicit acts, an atmosphere of nonchalance or dejection in which agents invest the (procedural) means of bureaucracy at the expense of its goals dominates many a government office.
Some well-functioning organizations do exist in dysfunctional environments. “Islands of competence,” as Geddes (Reference Geddes1996, 23) calls them, enable parts of the state to successfully carry out their duties (e.g., Bowman Reference Bowman2015). Working on Ghana, McDonnell (Reference McDonnell2017) refers to them, collectively, as an “interstitial bureaucracy” operating in the cracks of an otherwise ineffective state apparatus. But how do such “islands” come about? What explains successful reforms?
Scholarship on the state and institution building points to several factors. The structuralist strand of research uncovers internal and external pressures that compel rulers to create strong state institutions, but it glosses over the mechanics of actually building a state (Doner et al. Reference Doner, Ritchie and Slater2005). Scholars with a more institutionalist bent have identified other conditions, including political will (Geddes Reference Geddes1996), decentralization (Velasco Jaramillo Reference Velasco2014), gradualism (Bersch Reference Bersch2016), and compromise (Brinks et al. Reference Brinks, Levitsky and Victoria Murillo2019). Staff vetting and training, coupled with adequate pay, are also key to disincentivizing corruption. The Weberian ideal of tenure, autonomy, technical skill, and commitment still informs much of the thinking and writing on state building.
However, the disjuncture between the formal organization of bureaucracy and everyday practice means that implementing the “right” set of incentives does not guarantee success (Rueschemeyer and Evans Reference Rueschemeyer, Evans, Peter, Rueschemeyer and Skocpol1985). Institutional legacies, enforcement gaps, and the fact that new incentives play out over time, with unintended consequences, undermine state-building projects (Becker Reference Becker2024; Brinks et al. Reference Brinks, Levitsky and Victoria Murillo2019). Most studies acknowledge the underlying layer of organizational life—sometimes called “informal institutions” (Helmke and Levitsky Reference Helmke and Levitsky2006)—upon which reforms tend to “crash,” but its structure and dynamics have received less theoretical attention. Improving the rules of the game by increasing transparency or pouring more resources into enforcement efforts can go only so far in overturning the time-shaped soul of bureaucratic organizations.
Drawing on Selznick’s (Reference Selznick1957) definition of institutionalization as “infusing with value,” this article links the formal and informal dimensions of institution building. It shows that charismatic leadership can shift value commitments toward compliance and public service. Its empirical basis is the reform of Peru’s tax authority, Superintendencia Nacional de Aduanas y de Administración Tributaria, or SUNAT, which underwent an institutional makeover in the early 1990s. Several conditions identified in the literature as propitious for reform were met, including an economic crisis, a recent election, international pressure, presidential support, and a change in leadership. The new leadership implemented favorable measures, such as personnel vetting and pay increases. Yet the distinctive feature and the active ingredient were the images and performative actions of the people in charge. The charismatic performance of a nation-building enterprise by the newly appointed director Manuel Estela and his team gave credence and value to what seemed to many, at first, to be yet another hopeless reform attempt. The Weberian typology and Weber’s writings oppose bureaucratic and charismatic domination—the latter being more prevalent in premodern times, the former taking over through institutionalization processes as the “iron cage” of rationality expands (see Weber Reference Weber1968). And yet it was a charismatic force that paradoxically laid the basis for legal-rational domination at SUNAT. Studies of institution building, especially those in the new institutionalist vein, have so far paid little attention to personalized forces and stories (Mahoney and Thelen Reference Mahoney and Thelen2009; Peters Reference Peters2019; Scott Reference Scott2013). This case study examines the concatenation of the foregoing factors, with a focus on charisma, and teases out their influence on the shift toward values of public service at the agency. The discussion takes a comparative look at another tax agency reform in the region.
Reforming the state
A significant body of research frames the problem of building state capacity as a matter of government decisions. In this view, functional bureaucracies emerge only if and when rulers choose to prioritize the public good over private gain or short-term political returns. As Cárdenas (Reference Cárdenas2010, 3) puts it, “State capacity is the result of decisions taken by governments, which [face] different incentives and constraints.” In many cases, such incentives and constraints favor patronage, including the distribution of public sector jobs to party loyalists and related neo-patrimonial practices (Nichter Reference Nichter2018; Ordóñez Reference Ordóñez2024). According to Geddes (Reference Geddes1996), term limits can create a safeguard, as politicians not seeking reelection are less likely to use the state to secure votes and more likely to appoint competent technocrats at the helm of state agencies.
Other scholars highlight crisis and urgency as key drivers of state building (Gil Reference Gil2022). In their comparative study of the origins of the developmental state in Southeast Asia, Doner et al. (Reference Doner, Ritchie and Slater2005, 328) argue that “political elites will only build such institutional arrangements” when facing a risk of “unmanageable mass unrest,” a threat to national security, and “hard budget constraints.” Less dramatically, “crises provide leaders with extraordinary opportunities to demonstrate their capacity to lead and fulfill aims that would be impossible to achieve under normal circumstances” (Boin et al. Reference Boin, Stern and Sundelius2016, 4). The crisis thus opens the proverbial window of opportunity for politicians and technocratic agents to act.
And yet the growing new institutionalist literature reveals complexities in the workings of the state that hinder change even under such “favorable” circumstances (Steinmo and Thelen Reference Steinmo and Thelen1992). Institutional change is path dependent, and reformers face the deadweight of a constraining legacy, which limits the range of options available to them (Collier and Munck Reference Collier and Munck2022; Pierson Reference Pierson1995). Moreover, the institutional weaknesses that plague Latin American bureaucracies entail a discrepancy between the official organizational framework, on the one hand, and everyday administrative practice, on the other (Centeno and Portes Reference Centeno, Portes, Fernandez-Kelly and Shefner2006). In this context, the standard techniques of institutional engineering have limited effects and unintended consequences, regardless of which paradigm reformers choose to follow. There is an informal layer of bureaucratic life, and reforms disembedded from the social life of weak bureaucratic organizations are unlikely to deliver on their own.
A drastic response to this challenge involves replacing at least part of the staff. This path raises ethical concerns, however, and, in most countries, faces major legal hurdles. It is also double edged. Although part of the staff may be corrupt, another part—or even, tragically, the same part—may hold valuable expertise without which the agency’s operations would be compromised. And there is no guarantee that new recruits will follow the right path. It is one of sociology’s foundational insights, articulated by Durkheim ([Reference Durkheim1895] Reference Durkheim1988), that institutions exist beyond and, to a large extent, independently of the people they govern. The agency could convert new employees far more easily than they would reshape the organization. The difficulty is especially acute in the field of taxation, which stands among the top policy challenges for Latin American governments (Dewey Reference Dewey2018; Díaz-Cayeros Reference Díaz-Cayeros2006). The power of business interests, combined with individual opportunities for illicit enrichment, makes it all the more difficult to establish a culture of public service.
Time and values
Echoing Weber, scholars interested in the foundations of statehood often allude to a bureaucratic ethos. The term ethos suggests a system of normative dispositions—that is, values—that operates beyond pragmatic adaptation to, or self-interested abidance by, a set of formal rules. There is, in other words, a spirit of bureaucracy underlying its blueprint. McDonnell (Reference McDonnell2017) makes the case that such ethos can exist independently of formal structure and that some civil servants actually circumvent the rules to achieve public interest goals.
But where does the bureaucratic spirit come from? Old institutionalism points to socialization (Stinchcombe Reference Stinchcombe1997). Formative institutions endow their members with a set of values—which members internalize—and, in most cases, specialized skills. The army is the quintessential example of such value-inducing institutions, but states with Weberian bureaucracies tend to have similar mechanisms covering a wider population of civil servants, which build upon earlier agents of socialization—especially school—and instill the values of public service. France, for example, is known for its grandes écoles, which have trained multiple generations of top-ranking civil servants (Bourdieu Reference Bourdieu1998). Of course, perfect socialization exists only in theory, and even totalitarian regimes leave room for individual subjectivity. What matters here is that socialization is a long-term process of moral conditioning that cannot be enacted overnight. And in the absence of proper socialization to public service, the ideal of committed civil service carries little weight among public sector employees while self-interest trumps institutional motives. As Rueschemeyer and Evans (Reference Rueschemeyer, Evans, Peter, Rueschemeyer and Skocpol1985, 45) put it, “The existence of an adequate bureaucratic machinery depends on a more delicate, long-term process of institution-building.” Putnam (Reference Putnam1993, 60) echoes this statement when he notes that “the rhythms of institutional change are slow. Often several generations must pass through a new institution before its distinctive effects on culture and behavior become clear.” In a context of weak institutional foundations, then, to build state capacity is to go against history. As Weber (Reference Weber1947, 362) puts it, it implies “repudiat[ing] the past.”
Not any kind of leadership can accomplish such a feat. Public management studies emphasize organizational leaders’ role in defining a mission and stirring up motivation. Strong leadership fosters team spirit and cooperation, which helps attain institutional goals. However, leading the reform of a corruption-ridden administration involves not only steering the ship in a new direction but also uprooting a legacy of malpractice, with its attendant habits and baneful values. A new, committed leadership may be a necessary ingredient for turning a corrupt organization around, but it is not a sufficient condition. In the language of management, under these circumstances, leadership needs to be more than inspirational. In Selznick’s words, leadership has to “build [the values of public service] into the social body of the organization.” Not only assert the values; make them meaningful. Hence, reframing institutional values goes one step further than “the same old idea,” as Beyer (Reference Beyer1999, 309) puts it, “that certain kinds of leaders can change followers’ attitudes, beliefs, and behaviors, and in this way, induce improved organizational performance.” It involves changing what people want and what they think is possible. It is a disruptive process.
Charismatic agency
The concept of charismatic leadership, which Max Weber (Reference Weber1968) famously introduced into the social sciences, made a deeper mark in social psychology and management studies than in the discipline Weber contributed to founding. Part of the problem stems from skepticism among sociologists toward a concept that presumes outsize personal influence—or, worse, in its original sense, a gift from heaven. In fairness, the essence of charisma is elusive, and students of charisma have grappled with the conceptual problems it raises in different ways (Antonakis et al. Reference Antonakis, Bastardoz, Jacquart and Shamir2016).
Conventional wisdom ascribes charisma to an individual leader. Rare personality traits or a special mix thereof give exceptional individuals—such as Jesus or Churchill—exceptional sway over those around them. The idea that individuals have intrinsic, independent, and stable personality traits is a sociological nonstarter, however. Like most of his followers, Weber emphasizes perception. From this perspective, charismatic leaders are those whose qualities are perceived as extraordinary or superhuman by their followers. Meindl (Reference Meindl1995) goes so far as to suggest that followers construct the charismatic leader, under particular circumstances, almost independently of her actions and personality. What matters is that charisma is produced as part of a social relationship involving one or multiple groups and audiences.
Because historical leaders widely recognized as charismatic also seem to have had widely different personalities, the question remains as to what enables them to persuade others or what leads others to construe them as exceptional. Different authors cite different features, such as willingness to die (Turner Reference Turner2003) and radical vision (Trice and Beyer Reference Trice and Beyer1986). Andrews-Lee (Reference Andrews-Lee2021) emphasizes the leader’s recognition of suffering among the people and their promise to assuage it, which hints at the importance of narrative in the construction of charisma (see also Andrews-Lee Reference Andrews-Lee2019). The leader’s features, words, and actions acquire their significance in connection to the situation in which charismatic leadership arises, typically one of crisis. Crises call for saviors; they predispose the masses to give credence to someone who claims to know the way out of the current predicament and is willing to make the necessary sacrifices; by implication, they turn the leader into a unique and uniquely endowed human being. Some charismatic leaders emphasize the theme of crisis (Zúquete Reference Zúquete2011); virtually all of them spin a narrative of salvation (Zúquete Reference Zúquete2020).
Even then, determining who qualifies as charismatic is tricky. A would-be leader can check many boxes and yet fail to sway the masses, even in times of crisis. Or they can be perceived as exceptional by some people and not others. Friedland (Reference Friedland1964) thus uses “registered success” as a defining criterion. Because the recent uptake of interest in charisma—which dovetails with the global rise of populism—has focused on political actors, the (implicit) test of charisma is often winning elections or related indicators, such as the size of the vote, approval ratings, and public expressions of allegiance.
Charismatic discourse holds a bigger promise than just access to power, however. If leaders base their legitimacy claims on a savior narrative, do they effect substantial change for the better when empowered? Or is charisma precisely the ability to keep that support and enthusiasm alive, notwithstanding the absence of material improvements? Charismatic leaders seldom deliver the kind of revolutionary change they promise, and defining a concept on the basis of expected outcomes raises serious methodological problems (see Antonakis et al. Reference Antonakis, Bastardoz, Jacquart and Shamir2016). But the performative element by which the narrative comes to life through publicized acts of heroism—real or forged—plays an essential role in the making of charismatic authority, along with tangible, if often transient, results leaders can point to (Andrews-Lee Reference Andrews-Lee2021). Hence the centrality of signaling (Antonakis et al. Reference Antonakis, Bastardoz, Jacquart and Shamir2016; Grabo et al. Reference Grabo, Spisak and Van Vugt2017).
Drawing on the previous debate, I define charisma as a collective performance articulated around an individual figure with exceptional qualities in the eyes of her followers. What those qualities are matters less than the fact that they are perceived as exceptional. The charismatic leader puts forward not only a vision but also a narrative about the future, which can be nationalistic (de la Torre Reference de la Torre2013), theological (Turner Reference Turner2003), or transcendent in some other way. And the charismatic performance involves displaying steadfast adherence to the narrative in the face of adversity, as well as showing its partial realization, which broadens or strengthens adherence beyond the circle of hardcore believers. (The religious themes of prophecy, faith, and salvation are difficult to escape.)
I argue that charisma has transformative power—specifically, the power to change values. Some supporters may believe that redemption is around the corner, but others clearly follow the leader for what she or he stands for. In their case, support is not irrational. It signifies adherence to a different set of values. On account of the value they attach to the realization of the leader’s narrative and the faith they develop, followers are willing to make sacrifices—that is, to forgo material satisfaction in pursuit of a greater good, in line with a basic tenet of the bureaucratic ethos. This value shift closes the enforcement gap at the heart of “frustrated states” (Centeno and Portes Reference Centeno, Portes, Fernandez-Kelly and Shefner2006) and addresses institutional weakness, not by increasing enforcement capacity through surveillance or coercion, but by fostering spontaneous compliance and alignment with larger goals.
Methods
Fieldwork took place in Lima, where SUNAT’s headquarters are located, and in Quito, where I conducted interviews with four former employees of the Servicio de Rentas Internas (SRI) and a former cabinet minister. In Peru, I conducted twenty in-depth interviews and analyzed about 150 documents from SUNAT’s archives and other sources. The interviewees included current and former high- and middle-ranking government officials who participated in the reform, including three former directors of SUNAT and one former vice director. I also interviewed a historian, two economists, and a former congressman who took part in a vote over reform legislation.
From a broader methodological perspective, the reform of SUNAT considered in this study is a case of the least likely outcome in institution building. Many reforms had been launched before the one under consideration; all consistently failed. By helping to identify the distinctive forces at play in the successful instance, which defied both theoretical expectations and some of the actors’ predictions, the analysis provides theoretical insight into what contributes to effective reforms.
The focus on personal agency raises another question. Social scientists are rightly skeptical of accounts that place exceptional characters at the center of a story about major social change. In this regard, the following account pays particular attention to a range of systemic factors, including the economic crisis and external pressures, which enable the intervention of personal actors. At the same time, it is a (social) fact that exceptional personalities do exist and that, under special circumstances, they can play a significant role in shaping the course of events. Rather than dismissing the study of charisma as hagiographic, it is important to put the action of charismatic leaders in sociological perspective, as Weber did, and to show how their action is mediated by structures, groups, and the perceptions of those around them.
The institutional history
SUNAT was born in the waning days of the first García administration (1985–1990) as a desperate attempt to get public finances under control. The country was engulfed in chaos. Between February 1988 and May 1990, the Peruvian economy contracted by 23.4 percent (Weyland Reference Weyland2000, 485). Inflation reached unprecedented annual rates of 2,774 percent in 1989 and 7,335 percent in 1990. During the last four months of 1988, consumer prices soared by a staggering 14,900 percent (Sachs and Paredes Reference Sachs and Paredes1991). Under these circumstances, increasing tax revenues was not only desirable—it was vital. (Under García’s successor, it was part of a stabilization program that included spending cuts and restrictive monetary policy.) In June 1988, at the president’s behest, Congress voted to approve Law 24892, which established SUNAT as an “autonomous” tax agency with a “budget of its own.” The new administration’s prerogatives were restated in a legislative decree issued by the president in December 1988, which specified the organizational structure of the newly created agency.Footnote 1
On top of the economic breakdown, the social fabric was torn apart by a civil war pitting the state against the Maoist guerilla Shining Path, which had extended its operations from the south-central highlands to the capital city. John Crabtree (Reference Crabtree, Crabtree and Thomas1998, 17) described the social atmosphere of the time as “a psychosis of insecurity which undermined the credibility [of formal democratic institutions].” In this context, Alberto Fujimori, a complete outsider, defeated future Nobel Prize winner Mario Vargas Llosa in the April 1990 presidential runoff. During the twenty-one months that followed Fujimori’s inauguration, 1,214 terrorist attacks took place, killing more than two thousand people (Bowen Reference Bowen2000, 100).
Moreover, the formal steps of creating a new tax agency had done little to improve the state’s finances, as the new agency was still plagued by the same problems that undermined its predecessors (Arias Reference Arias1995; Durand and Thorp Reference Durand, Thorp, Crabtree and Thomas1998; Zapata Reference Zapata, Contreras, Mazzeo and Quiroz2010). Hence, in 1991, the recently elected Fujimori launched a tax administration reform and appointed Manuel Estela, an economist from the Banco Central de Reserva (BCR), for the job. As opposed to all previous attempts at reform, the reform of SUNAT under Estela and his team resulted in a total revamping of the organization based on a new and surprisingly efficient public management model. Change is reflected in the ratio of collected taxes to gross domestic product, which grew for four consecutive years following the launching of the reform in 1991. In 1990, 700 taxpayers accounted for 90 percent of tax revenue (Arias Reference Arias1994, 2). By July 1993, there were already 558,000 registered taxpayers. Moreover, the number of high-income taxpayers—known as pricos (principales contribuyentes)—went from 1,212 to 7,400, and by 1994, they contributed 63 percent of tax revenue (Arias Reference Arias1995, 7).
This dramatic upward turn translated into public opinion polls, which cast SUNAT’s reform as the third-biggest achievement of the Fujimori administration (Durand and Thorp Reference Durand, Thorp, Crabtree and Thomas1998). Along with the church and the army, SUNAT stood as one of the most trusted public institutions in the country. And the culture and climate of the organization underwent a revolutionary transformation: Selfless effort and service to the public became ingrained values of administrative work. What brought about such fundamental change in an ineffective and informally run organization that was overtaken by corruption and powerful mafia-like cliques?
The institutionalist explanation posits that the social, political, and economic crisis—coupled with recent elections, international pressures, and political will at the top—formed a critical juncture that broke the previous development path predicated upon graft, patronage, and political interference. The Peruvian sociologist Francisco Durand, who participated in the reform, acknowledges these factors while highlighting the cohesiveness, competence, and integrity of the reform task force (Durand and Thorp Reference Durand, Thorp, Crabtree and Thomas1998). In what follows, I discuss the interaction between long-term institutional dynamics and the critical juncture in SUNAT’s reform before considering the role of charismatic leadership.
A corrupting path
The roots of SUNAT’s problems ran deep. In 1931, Peru played host to the Kemmerer Mission, an official US government mission named after its head, Edwin Kemmerer, a professor of economics at Princeton. Inspired by the US system, the Kemmerer Mission recommended reinforcing the independence of the BCR, which had been established ten years earlier under Leguía, as well as creating an autonomous internal revenue agency. As recounted by Zapata (Reference Zapata, Contreras, Mazzeo and Quiroz2010), the 1933 constitution enshrined the independence of the central bank, in line with Kemmerer’s advice, while tax collection remained in the hands of a private banking entity, the Caja de Dépositos y Consignaciones. A small administrative unit within the Ministry of the Economy and Finances (MEF)—then called Ministerio de Hacienda—exercised oversight.
As a result, the central bank and the tax authority followed divergent institutional paths. Drawing on its constitutionally embedded autonomy, the BCR developed an institutional policy of personnel management based on expertise and seniority. It introduced a competitive entrance examination in the 1950s. By the 1980s, applicants were mostly high-performing university graduates with degrees in economics. To join the bank staff, they had to take part in a three-month-long training program held each year during the summer, at the end of which they took the entrance test. Only the five best-scoring candidates were hired—and given tenure. Using its own funds, the bank would send its employees to pursue master’s and PhD programs at prestigious universities in the United States and Europe.
By contrast, tax administration did not undergo any institutionalization process. Tax collection was brought under the helm of the state only in 1966. Even then, it remained a department within the Ministry of the Economy, under the executive’s direct control (Zapata Reference Zapata, Contreras, Mazzeo and Quiroz2010). It achieved autonomy—on paper—only in 1988, with the creation of SUNAT. SUNAT’s autonomy thus came more than half a century after the BCR’s, and it was not written into the constitution.
Tax administration in Peru thus followed what might be called an informal path of institutional development. Even though the organizational framework proclaimed basic principles of bureaucratic administration, these were loosely applied, and sometimes blatantly transgressed, in the absence of underlying social values of public service. In lieu of a bureaucratic ethos reinforcing itself over time by means of in-group transmission and customary practice, an informal culture of job rigging, bribe taking, rent seeking, and coercive hoarding of public resources became entrenched.
Unsurprisingly, SUNAT inherited these problems upon its creation. Some workers were being investigated for their presumed involvement in a multimillion-dollar tax evasion scheme.Footnote 2 Some services had fallen prey to mafia-like cliques who used intimidation and coercion, sometimes engaging in taxpayer extortion (Durand Reference Durand, Thorp, Crabtree and Thomas1998). Beyond the most egregious acts, large-scale, discrete, and routine corruption practices plagued administrative life, according to various accounts. The unlawful provision of services to private firms by tax auditors, for which they received monetary compensations, was not uncommon.Footnote 3 Francisco Durand (2002, 17) documented other illicit transactions, like the slippage of banknotes in the folds of administrative paperwork.
In fairness, this state of affairs fits the general picture of the Peruvian state at the time (Contreras and Cueto Reference Contreras and Cueto2007). Against this background, the measures taken to prop up tax administration under García remained dead letter. The new agency was overridden by the general crisis and unable to enforce its autonomy against the MEF. It also lacked a qualified body of bureaucrats committed to the task. To make things worse, on his way out, García indulged in some old-fashioned patronage. In June 1990, SUNAT workers staged a strike protesting the hiring of “hundreds” of apristas upon presentation of their party credentials.Footnote 4
In this gloomy landscape, the BCR—whose board eventually caved to García’s pressures—still boasted a reputation of technical expertise and professionalism. Its organizational model inspired the team of SUNAT reformers, the majority of whom were drawn from the bank, to which they frequently referred in interviews as their common “school.” To paraphrase Selznick (Reference Selznick1957), this amounted to a “transfusion” of values. On the other hand, few members of the “the BCR team”—as they often identified—had any expertise in taxes, except for Luis Arias, a young BCR economist with a master’s degree from the London School of Economics whom Estela appointed vice director of SUNAT.
The critical juncture
Not only was Fujimori a complete outsider when he won the 1990 presidential elections; he did not have a real party, ideology, or social basis that bound him to a specific policy. The new administration’s willingness to restore relations with the International Monetary Fund (IMF) required balancing public sector accounts and providing warrants of future solvency. In fact, the reform of the legal tax system, the increase of tax revenues, and public sector restructuring along new public management lines were constitutive items of the IMF-sponsored structural adjustment agenda, which Fujimori fully embraced, violating his campaign promise of gradualism. Under these circumstances, Fujimori had both a fiscal urgency and a strategic interest in supporting the tax administration reform and boosting tax revenues. Efficient reform rather than patronage politics—which he later lavishly practiced—was the rational response to Geddes’s (Reference Geddes1996) “politician’s dilemma” at that point. According to Durand and Thorp (Reference Durand, Thorp, Crabtree and Thomas1998, 215), when he appointed Estela, Fujimori pledged to get “personally involved” in the enterprise and emphasized his independence from powerful pressure groups whose interests the reform project was bound to run against.
The crisis did more than bring an outsider to power and create the incentives for serious reform. It enabled Fujimori to push through institutional change. As Keeler (Reference Keeler1993) notes, crisis situations are instrumental to the adoption of far-reaching reform legislation, for they create a sense of fear and urgency that weakens institutional resistance to, and bolsters popular support for, bold initiatives on the part of the executive. In Peru, normal legislative procedures were initially pressed, then disrupted. Under García and Fujimori, Congress granted the executive special legislative prerogatives. In the exercise of such powers, Fujimori issued “waves” (oleadas) of legislative decrees (Degregori Reference Degregori2000, 32). Some of these decrees directly concerned the implementation of the tax administration reform. A landmark text was Decree Law 639, issued on May 16, 1991, which launched the “reorganization” process and laid the legal basis for a thorough staff evaluation. By the same token, it empowered the new administration to investigate evidence of corruption and dismiss any civil servants displaying “ostensible” signs of having engaged in illicit transactions.Footnote 5 Four months later, Decree Law 673 granted SUNAT a special legal framework for personnel management modeled after private sector regulations.Footnote 6
In the political arena, the disruptive trend reached its peak in the autogolpe (self-coup) of April 5, 1992, when Fujimori, backed by the chairman of the Joint Chiefs of Staff, suspended the constitution and ordered the shutdown of both legislative houses. From then on, Fujimori governed single-handedly, with the constituent congress elected in October 1992 acting as a rubber stamp (Kay Reference Kay1996). In a view echoed by other reform actors, a member of Estela’s team claimed that during the time SUNAT still enjoyed presidential protection, the coup “made things easier” because the adoption of tax legislation no longer required political bargaining.Footnote 7 Critical tax policy measures were adopted during this period, like the compulsory use of customer receipts (comprobantes de pago), the creation of a centralized taxpayer identification system, the enactment of a tax code (código tributario), and the drastic simplification of the tax system, which brought down the number of taxes from more than two hundred to six.
Presidential support proved crucial for the reform in yet another way. In a regime where mediating and representative political institutions such as parties or congress fell into popular discredit and political ineffectiveness, public polls and electoral results became the fulcrum of a highly personalistic form of political authority (Degregori Reference Degregori2000; Kay Reference Kay1996; Weyland Reference Weyland2000). A share of his crisis-derived political capital was invested in SUNAT in the form of presidential appearances at spectacular events such as the operativos de fiscalización (tax campaign operations) conducted by the tax agency at informal marketplaces. On another occasion, the president attended a press conference at which SUNAT announced a crackdown on big tax evaders and praised in his speech the “extraordinary and relentless [tesonero]” labor of Estela and his team.Footnote 8 Fujimori also made use of his authority behind the scenes, intervening to curb attempts by one of his ministers to undermine the prerogatives of the autonomous tax agency by limiting its financial resources in a draft of the 1992 budget law.Footnote 9
From a new institutionalist perspective, institutional change at SUNAT was possible—even likely—at the time. Critical junctures are “moments of heightened uncertainty” in which the usual constraints that shape social and political life lift as a result of exogenous shocks or incremental processes, and trajectories that were previously unthinkable become available (Capoccia and Kelemen Reference Capoccia and Daniel Kelemen2007; Collier and Munck Reference Collier and Munck2022; Mahoney and Thelen Reference Mahoney and Thelen2009). Some accounts of critical junctures are more structural in that the main force is the timing of events over which individual actors have little leverage (Collier and Collier Reference Collier and Collier1991). Others make room for political agency, especially on the part of the president and/or the head of the executive branch (Collier and Munck Reference Collier and Munck2022; Hacker Reference Hacker1998; Pierson Reference Pierson1995). Indeed, power and decision-making are part of new institutionalist thinking, especially insofar as institutions shape attributions and incentives. As noted earlier, however, institutions have multiple layers, and reforming them requires more work than legislative change, behind-the-scenes political backing, and the media support that Fujimori provided. Personal and symbolic elements need to be factored in for us to understand what happened inside the tax agency.
The personal touch
When Alberto Fujimori appointed Manuel Estela director of SUNAT in March 1991, his mission was to reform the tax authority. The choice of Manuel Estela for the job was reportedly suggested to the president by his brother, Santiago Fujimori. Santiago Fujimori and Manuel Estela were almost the same age and had known each other from childhood.
Estela was a former seminarian who had given up clerical life and joined the BCR in his youth. He had graduate training in economics from an elite UK university. Interviewees described him as “special” or even “weird,” and several of them attributed his peculiarity to his “theological background” or “spiritual formation.” Estela did not have the flamboyant, rabble-rousing personality that fits a common stereotype of charismatic leadership. In line with another popular cliché, however, he embodied the righteous and austere guide.
As a career civil servant at the BCR, Estela had few incentives to accept the daunting task of building a functioning tax agency. In his telling, when he met with the president, Estela set three conditions: autonomous decision-making, the ability to build a team of collaborators of his choosing drawn from any public institution, and the authority to turn down demands for favors from congresspeople.Footnote 10 The president agreed. Estela picked a task force drawn mostly, but not exclusively, from the BCR.
A nation-building enterprise
Estela had a clear view of the nation and its enemies. In the book El Perú y la tributación, he rails against the “pseudoliberals”: those members of the economic elite who, camped behind free-market rhetoric, plunder the resources of the state and the nation. The book’s opening cites the story of the famous hacienda La Brea y Pariñas, which was sold for a pittance to a foreign oil company at the beginning of the twentieth century. It suggests a parallel with the action of powerful business groups who attempted to curb the reformers’ efforts to enhance the autonomy and effectiveness of the tax authority. Both cases, in the author’s view, illustrate a lack of commitment to the national interest by the ruling elites.
To achieve “sustained growth and welfare,” Peru had to overcome two obstacles: a “chronic fiscal [imbalance]” and an “improvised state.” “Only a neutral and fair fiscal system,” Estela (Reference Estela2002, 39) writes, “will allow us to fund in a salutary manner the redistributive policies that will lift from their condition of marginality and poverty those 50 percent of the population barred from the benefits of the market.” At stake in SUNAT’s reform, then, was the fate of the nation. And Estela managed to convince at least some of the staff.
As career civil servants with a background in economics or public finance, task force members and SUNAT employees clearly understood the link between the collection of taxes and the severe fiscal and economic problems facing the country. Paulino Barragán recalls how, in those days, civil servants were paid with sheer “paper” because of the complete lack of “liquidities” in state coffers. Another interviewee points to the government’s incapacity, under the circumstances, to simply honor its payroll. Estela’s charismatic performance managed to raise the stakes, however. According to Luis Francia, a former Banco de la Nación employee who joined the team: “[The reform endeavor] was a commitment with the country. We worked to raise tax income with an exclusive goal. That exclusive goal was to step out of underdevelopment or, in more concrete terms, to solve the problem of poverty.”Footnote 11 A young human resources manager who joined SUNAT in the mid-1990s echoes this sentiment: “A lot was expected from you. The nation was driving you [el país te movía].”Footnote 12
This nation-saving effort translated into a workplace mystique (“una mística”). Working hours extended far beyond contractual obligations, overnight and weekend stays at the office were not uncommon, and taking holiday leaves was frowned upon by colleagues. A mid-level employee said he was “taken [contagiado] by this group of compañeros led by Estela.” In the poetic words of Luís Francia, SUNAT at the time had a “personal charm” (un encanto personal). The leader himself was keenly aware of these dynamics and their importance. In a specialized magazine article where he reviews the key factors of SUNAT’s initial success, Estela (Reference Estela2000, 39) points to the “sharing of a transcendental sense attributed to the levy of taxes: each [employee] from their position and by their daily activity felt they were contributing to the construction of hope for Peru.” The aim was not only to “modernize” tax administration or to raise tax revenue but also to “build a nation” (construir país).Footnote 13
How did Estela convey these meanings? Inquired about the drivers of his unrestrained investment in the project, one member of the “BCR team” points to Estela’s command of magic “palabritas” (little words). Several expressions attributed to the leader, like the metaphor that achieving the reform was akin to “crossing a river” or that, in spite of discouraging antecedents, the reform would succeed because “planets [were] aligned,” were used by other interviewees to frame their own accounts of the experience. Quite literally in these examples, leadership provides meaning. Cues and signals are at the core of Grabo et al.’s (Reference Grabo, Spisak and Van Vugt2017) conceptualization of charisma, and Antonakis et al. (Reference Antonakis, Bastardoz, Jacquart and Shamir2016) allude to the leader’s command of metaphors.
Documents and rituals backed up speech in the performative construction of the leadership. From the outset of Estela’s administration, a “decalogue” was published that stated the norms of organizational life. Alcohol was banned from any kind of institutional celebration, and auditors were forbidden from accepting the slightest gift from taxpayers, including, in one account, a glass of water. The leader’s status was also constructed in daily acts, such as returning personal gifts from private actors—apparently a common practice before the reform—in a conspicuous manner, “for everyone to see.”Footnote 14
But charisma cannot stand on routine practices alone, however admirable. Its narrative is too far-fetched, too much at odds with everyday life and the (cynical) common sense that derives from it for it to sway more than a small nucleus of loyal followers through ordinary means. In the case of SUNAT, dramatic episodes contributed to cementing the perception of the leader’s exceptional qualities and gave credence to his claims. Thus, Paulino Barragán, a senior employee, recounts the aftermaths of the bombing of SUNAT’s headquarters in March 1992: “The leader immediately appears on TV and says: ‘Ladies and gentlemen, what just happened only strengthens our conviction that we are on the right track. [As the tax administration gets to work], there will be no more such attacks because we will be able to address people’s hunger, their basic needs, et cetera.’” Barragán quotes a line from Estela in that televised address: “Because I will not respond to bullets with bullets, but with work,” and concludes, “That reinforced our convictions of service to the nation.”Footnote 15
The ultimate test of the charismatic performance, however, is reality itself—or, as Weber (Reference Weber1978) puts it, “corroboration.” Emotional statements aside, the leader must show her powers. The bombing, which took place on a Friday, offered an opportunity to do so. By Monday, the staff was back to work. As Barragán puts it, “El hombre habla y se cumple” (What the man says gets done). This contrasted with the empty promises of change to which SUNAT workers and the Peruvian public in general had become accustomed. Estela became a popular public figure beyond the organization.
Of course, not everyone at SUNAT was equally imbued with the nation-saving narrative and its attendant values. Some interviewees mentioned the lust for challenge or promotion opportunities as their reasons to get involved while also alluding to a disposition for public service. Despite generating the mystique, moreover, charisma does not work miracles on its own. Personnel selection and institutional incentives also matter, as reformers understood well.
The selection process
Though barely implemented, the decrees issued under García had laid the formal organizational framework upon which reformers would build. However, the subsequent Decree Law 639, whose draft was written by members of the reform team, asserted the institutional project’s founding values: “a solid and permanent institution, in which the exercise of civil service should be technically and morally exemplary.”Footnote 16 By the same token, it established the legal framework to undertake a thorough personnel examination and selection.
On the eve of the reform, SUNAT had 3,025 employees, two-thirds of whom were technicians and clerical workers (técnicos) and one-third of whom were professionals (profesionales). Those believed to hold valuable in-house expertise were individually targeted and persuaded to remain in the organization.Footnote 17 Then, except for a few cargos de confianza—positions “of trust” where the director could personally appoint collaborators—all offices were made vacant, and the selection process was launched anew. SUNAT employees were free to apply for the position of their choice. Those who refused to undergo the selection process were given the alternative of “voluntary resignation” under a special incentives program. Employees had a ten-day window to make their decision.
The plan came close to failing. Two days before the deadline, workers remained expectant while unions called for a boycott. In the end, contingent incentives (e.g., a fifty-dollar bonus for those who took the test), extensive dialogue between reformers and employees, and in one account, the appraisal of the reformers’ moral determination, spurred cooperation to snowball.
Appointees to each position were then selected by scrutinizing the candidate’s professional, technical, and “ethical” credentials. All applicants had to take a multilevel test, which probed technical knowledge as well as psychosocial characteristics. Although key informants dismissed the official story of a “test of ethics” as a subterfuge to scare bad apples, a whistle-blowing mechanism was put in place to identify corrupt or ethically dubious agents. In addition, the team conducted probes at the migration and property register offices. Individuals found to own goods or lead lifestyles out of proportion with their official income were dismissed and forbidden from reentering SUNAT under any circumstance. Luis Francia thus defined the initial evaluation steps as a search for people “who had not fallen in the bed of corruption.” In the end, only 991 out of the more than three thousand employees were retained, and the ratio of professionals to clerical or technician occupations was reversed. Drawing on Decree Law 673, reformers also introduced bonuses and raised wages to “competitive” levels. Salaries increased from an average of 50 USD in 1990 to 600 USD in 1991 to almost 1,000 USD in 1992.
At the end of the reevaluation process, more hands and brains were needed. An aggressive campaign to expand the tax basis geographically and across socioeconomic categories fueled the quest for more talented and committed professionals. The reform team came up with a new and quite peculiar hiring policy. One of its crafters recalled asking himself, “Where are those fools [sonsos] who still believe they can change the world?”Footnote 18 Only candidates younger than twenty-eight who had graduated from one of three elite universities in Lima were admitted to take the entrance test. Applicants who passed the test underwent extensive interviews in which top-ranking officials frequently sat on the committee, which was looking for “potential and attitude to undertake the project.”Footnote 19
In this context, SUNAT became a beacon for ambitious young graduates for its dynamism, career opportunities, appealing salaries, and positive public image.Footnote 20 Yet selection criteria were not relaxed. Of 9,340 job candidates in 1993, only 341 were hired, all of whom underwent the examination and training processes described earlier.Footnote 21 Those recruited would attend a compulsory five-week training program (the Curso de Administración Tributaria) and remain on probation for six months. Another peculiarity of the new organizational blueprint was the “ethical strengthening” seminars that all employees were required to attend.
It is not as though the reformers had a clear plan from the outset. They are the first to acknowledge how much they improvised as they went. One of the reform participants theorized their approach ex post facto, however. In a paper presented in Lima in 1995, Juan Carlos Cortázar (Reference Cortázar1995) characterizes the situation of the prereform tax agency as one in which the stated values and norms had become “purely formal.” To tackle the “irreducible exposure” of administrative agents to bribery solicitations and embezzlement opportunities, the author underscores the need to double the search for expertise with an assessment and assertion of moral values. Probity is also a recurring theme in Estela’s writings and the discourse of other reform participants.
In short, replacing values is a challenging task. In the case of SUNAT, it came with a heavy price tag. Around 2.3 million USD were distributed in the form of incentives and compensations. It also sparked violence. On top of the multiple boycotts, SUNAT was the target of two terrorist attacks during the reform implementation: a rocket aimed at the director’s office and a parcel bomb that exploded in the basement of SUNAT’s new headquarters in March 1992.
As noted, these incidents and the team’s reactions to them contributed to the construction of leadership. But interestingly, so did the new management policies. In the words of Jorge Napurí: “When you see that your leader gives everything [se mata] for the institution, when you see that he provides you with comfort, when you see that he gives you much better pay, when you see that he gives you better benefits with high-quality training … I’d kill for that institution.”Footnote 22
Shady aftermath
Estela resigned from office in November 1992. In an interview, the former director insisted that no one pushed him out and that he resigned when he realized that he had become too attached to the institution. There were also rumors that the president started to resent Estela’s popularity.
After Estela left, many of the people he brought remained at SUNAT. The lawyer who replaced him as director, Sandro Fuentes, was a member of the reform team. As the Fujimori regime grew more authoritarian and corrupt, however, special interests within the state began to thwart SUNAT’s autonomy. Fuentes’s successor, Adrián Revilla, struggled to maintain independence. Fujimori himself seems to have lost interest in protecting the agency as fiscal constraints relaxed and he prioritized his minister’s requests over Revilla’s. Later, during the second half of the 1990s, some SUNAT officials fell into the clutches of Vladimiro Montesinos, Fujimori’s nefarious security adviser who pulled the strings of much of the state apparatus by means of extortion, corruption, and crime.
Montesinos’s cooptation of SUNAT partly drew on the weakness of the director who succeeded Revilla: Jorge Baca Campodónico. Unlike Estela, who saw the MEF as an enemy, Baca Campodónico was close to the minister, Jorge Camet. After Montesinos approached Baca Campodónico with illicit requests, neither the director nor the minister could curb the powerful consejero. Montesinos co-opted another employee at SUNAT, Mauricio Muñoz-Nájar, who went on to become chief legal officer and then vice director thanks to pressure from another minister of economy. Both Baca Campodónico and Muñoz-Nájar were later indicted and received jail sentences. As for the members of Estela’s original reform team, several of them stayed at SUNAT through these years. In an interview, one of them said that, while she still controlled what happened in her unit, she had lost the power to shape the institution.
Discussion
In a nutshell, the institutional trajectory of SUNAT shows that critical junctures and executive backing are necessary but not sufficient conditions to effect institutional change within a government body; that charismatic agency has the power to change values and, in so doing, reverse institutional paths; and that charisma is a performance centered on an individual personality but reliant on a transcendent narrative as well as, in this case, a group surrounding the leader that conveys and performs the leading narrative. (In fact, when Estela showed that he was human and considered resigning after an attempt on his life, they convinced him not to.) By their statements and actions, the reform’s leading actors managed to enforce the belief that the reform was possible, that public order and social justice were at stake, and that strict compliance with institutional norms was both a value in itself and a necessary step to achieve the reform’s larger goals.
Katherine Bersch (Reference Bersch2016) distinguishes problem-solving, which is incremental, from “powering,” which is confrontational. SUNAT is an (extreme) case of powering. It was a conscious choice. Expressing a spontaneous understanding of path dependence, Estela noted that change had to happen all at once because, in Peru, “provisional things become permanent” (Durand Reference Durand2002, 20). Instead of “theorizing the institutional project in a way that resonates with the interests, values, and problems of potential allies,” as Battilana et al. (Reference Battilana, Leca and Boxenbaum2009, 80) suggest, Estela and his team put forward their own narrative of national redemption. Instead of seeking compromise with actors having a vested interest in the status quo, the reformers confronted them head-on and sought to wipe them out.
Bersch (Reference Bersch2016) argues that reforms based on powering are subject to abrupt endings when, for example, the head of government loses elections and the leadership is replaced. In the case of SUNAT, it did not take political alternation for the strategy to show its limits. A change in priorities and power relations within the executive branch thwarted the movement. Is charismatic state building thus bound to be short-lived? And what, then, is the significance of SUNAT’s experience in the history of the Peruvian state? After all, public sector institutions in Peru are still mired in corruption and other problems that reformers sought to address.
To paraphrase Geddes (Reference Geddes1996, 23), a pessimistic view of SUNAT’s reform would see it as a “flash of competence.” So long as a government agency operates in a political environment dominated by special interests, long-term change is unlikely. Nor is authoritarianism the answer. SUNAT’s autonomy initially benefited from the government’s autocratic turn, but then, as the regime became increasingly corrupt and abusive, so did SUNAT. National politics aside, some specific features of the reform, like the massive recruitment of young, bright, and ambitious workers at the outset, created problems down the road, as many faced a shortage of promotion opportunities and those dispatched to the provinces bemoaned the lack of opportunities to return to the capital city. But that is not to say the reform left no legacy.
For one, several members of the original team remained at the agency and continued to do their work dutifully, despite the challenges and conflicts that the leadership changes brought. In 2011, twenty years after launching the reform, Estela ran (unsuccessfully) for Congress. In televised interviews, he described SUNAT’s story as “tormented and painful” (atribulada y dolorosa). He bemoaned political interference and poor leadership. Yet he defended the cadre of “technically and ethically competent” employees SUNAT still boasted from the time of the reform.Footnote 23
Moreover, SUNAT’s experience gained international attention and inspired other tax administration reforms, such as in Ecuador, where the SRI was founded toward the end of 1997. The process benefited from the input of international consultants familiar with SUNAT’s experience. Like SUNAT, the SRI revamped its staff and targeted young graduates from Ecuador’s top universities. Like SUNAT, it offered competitive salaries on par with those of the private sector. And like SUNAT, the SRI started with an exceptional personality at its helm, Elsa De Mena, who infused the tax agency with a strong work ethic. “No dormíamos” (“We didn’t sleep”), said a former employee who joined the SRI in 1999. As did Estela, De Mena used little phrases that imprinted on her entourage; they speak to the signaling dimension of leadership.
Whether or not De Mena’s leadership qualifies as charismatic is a matter of definition. Her background and style differed sharply from Estela’s. De Mena hailed from one of Quito’s traditional families. She cut an iron lady figure, fearless and frightening, unafraid to swear or raise her voice. Hers was a no-nonsense, get-things-done approach. She ran the SRI like a corporation, with quantitative performance indicators and targets. Like Estela, De Mena had little phrases imprinted on her entourage that spoke to the signaling dimension of leadership. “Tengo que ser y parecer,” she used to say, or “Como te ven, te tratan.” Unlike Estela, however, she did not put forward a transcendental narrative about saving the nation, though fighting the ills of poverty was part of her pitch.
If the bar for charisma is set at eliciting adherence to a utopian project, then De Mena’s leadership style might not make the cut. There certainly was a dynamic of collective investment in the task at hand. Some young recruits “gave their hearts” and risked their lives to do the job. But their involvement does not seem to have had the same spiritual quality as SUNAT’s mystique. “Out of fear or out of conviction, people worked hard,” a top-ranking official under De Mena said in an interview. “And those who had neither did it for the money.”Footnote 24 Moreover, De Mena definitely had a special aura. In short, the comparison bears out that it takes exceptional personalities to revamp stagnant bureaucratic bodies.
The comparison with Ecuador is also enlightening with regard to the legacy. De Mena stayed on the job far longer than Estela—she ran the SRI for six years under four different presidents. (A former employee described her departure as an “earthquake” that created a state of “total consternation” among people who “didn’t know what to do.”)Footnote 25 When Rafael Correa succeeded Palacio as president in January 2007, he appointed Carlos Marx Carrasco as head of the SRI. By most accounts, Carrasco was not, in any way, a charismatic figure. And yet tax collection did not decline significantly on Carrasco’s watch. In fact, it continued its upward trend from 2007 to 2014—the year Carrasco left the SRI (see Vollenweider Reference Vollenweider2016).
A detailed comparison of SUNAT’s and the SRI’s experiences falls beyond the scope of this study. But Carrasco’s relative success suggests stability at the top is important—more so, perhaps, than continuity in terms of the managerial approach or vision. Carrasco was a Marxist political economist by training, working for a self-proclaimed left-wing populist president, whereas De Mena was an orthodox economist serving under Mahuad, a neoliberal president. But Carrasco seems also to have been committed to the task; someone who knew him well described him as “serious and responsible.”Footnote 26 While Carrasco brought in a lot of people, according to another interviewee, his administration did not kick anyone out. In time, however, factions formed, some of De Mena’s recruits sided with the new administration, the “camaraderie” dissipated, and some high-level officials engaged in dubious practices (e.g., arbitrary dismissals, sanctioning employees without due process).Footnote 27 And yet, even though he failed to instill the cohesive sense of mission that Estela or De Mena brought about, Carrasco’s long tenure still managed to build on De Mena’s legacy in a way that Estela’s short-lived successors did not. It helped, of course, that Correa’s government was committed to raising taxes. (Increasing direct taxes was a key plank of the ruling party’s program.)
From a more theoretical standpoint, the disappointing denouement at SUNAT raises Weber’s famous question about the routinization of charisma, which entails a construction of the leadership in the absence of the leader, through symbols and rituals. It means transferring the source of commitment and compliance to the organization itself—that is, employees feeling compelled to follow the rules and work toward the common good because they are members of SUNAT, not because they are caught in the mystique of a charismatic leader or the no-nonsense, get-things-done ethos of an imposing figure. This case study suggests that, if doable, such transfer requires that successors to the leader be steady-handed, if not charismatic, and that they continue to enjoy presidential protection.
Acknowledgments
I am grateful to Robert Jansen, Isaac Martin, Ari Adut, Michael Young, Philippe Bezès, and especially Yves Surel for their feedback on previous iterations of this project. I would also like to thank Abelardo Sánchez León and Antonio Zapata for their support and guidance during fieldwork. Part of this research was presented at the regular session on comparative/historical sociology at an annual meeting of the American Sociological Association in Las Vegas.
Open access
