Brasília was built at the end of the 1950s in the remote and previously almost uninhabited highlands of Brazil. It is shaped like a bird or an airplane, with two gently curved wings intersecting a long fuselage. The plan of the city was the vision of a handful of modernist architects, and to its different parts were assigned distinct functions. The main subdivision is between the residential wings and a long central body, where the political and administrative buildings are located.

I was in Brasília while researching this book on October 28, 2018, the day of the second round of the presidential elections that were to be won by Jair Bolsonaro. My hotel was located in the aptly named “hotel sector,” near the crossing of the residential wing with the governmental section. Brasília is a city envisioned for motor vehicles; there were not many pedestrians around me when, as the sun was setting on a clear day, I headed toward the tip of the fuselage, where the most important political institutions are located. I walked across the vast esplanade from which the cathedral emerges as a flower, and I soaked into the vision of that and other modernist buildings by Oscar Niemeyer, who designed much of Brasília’s iconic architecture. I saw a sad-looking girl, draped with a red flag, walking away alone, I thought, from some last-ditch political event. On the flag, written in block letters, I could read the name of Haddad, the candidate of the Partido dos Trabalhadores (Workers’ Party, or PT) and the loser of the day.

I walked past a string of ministries, most of them built alike, and finally reached the political heart of the city and the National Congress. Slowly, Bolsonaro’s fans started to assemble for what unanimous opinion polls had announced would be a historic victory. There were vendors selling Brazilian flags and assorted Bolsonaro merchandise, a few food trucks, and a truck with loudspeakers blasting out pro-Bolsonaro songs. One person, wearing a black robe with a hood and holding a long sickle in his hands, was available for anyone who wanted to take a selfie with Death. More opportunities for selfies became available once a large cardboard cutout of a smiling Bolsonaro started making the rounds in front of the esplanade facing the parliament building. A honking motorcade started soon after the results of the elections were made public, only hours after the polls closed, thanks to Brazil’s modern voting infrastructure.

The election of Jair Bolsonaro as president of Brazil followed a series of major corruption scandals, collectively known as “Lava Jato,” or “Car Wash,” which engulfed the PT and other political parties. Lava Jato uprooted Brazil’s political life and resulted in the impeachment of its serving president, Dilma Rousseff, and the imprisonment of her predecessor, Luiz Inácio Lula da Silva. The widespread popular mobilization of Brazilians opposing corruption ultimately enabled the election of an unlikely candidate: Jair Bolsonaro, a far-right, long-term member of parliament known for his sexist and homophobic remarks, who defended Brazil’s dictatorship, which ended in 1985, torture, and killings by a police force known for its abuses.

Brazil is a great and fascinating country that four decades ago managed a peaceful and largely successful transition from dictatorship to democracy, though certainly while displaying contradictions. The presence of pervasive corruption has been one of them, but democratic Brazil also carried out a long string of anti-corruption reforms. Ironically, they contributed to Lava Jato itself and placed at center stage a judiciary that would also be embroiled in controversy. Following a major leak of private conversations by The Intercept Brazil, an online news organizations, it became apparent that the heroes of the anti-corruption drive, Sérgio Moro (then a federal judge) and Deltan Dallagnol (then a federal public minister), had been involved in collusive practices that at a bare minimum were very inappropriate. Moro’s decision to accept the offer from newly elected President Jair Bolsonaro to become minister of justice further convinced many that justice had been weaponized to serve partisan interests. The raucous public debate that followed the Lava Jato case also makes Brazil a paradigmatic example.

Brazil’s anti-corruption reforms were not perfect, but to the extent that public policy is, of necessity, “muddling-through,”Footnote ² they were more ambitious than many other corruption-ridden countries could realistically hope for. They also aligned with the principles of the prevalent anti-corruption playbook. Civic society involvement, in particular, was nothing short of spectacular in the Brazilian case, if we are to judge from the massive popular mobilization in support of Lava Jato. When considering Brazil’s failure, as with all medicines that are not efficacious in a particular case, it might be argued that the doctor was inept or the patient recalcitrant. However, the lack of successful anti-corruption policies globally suggests that the problem may not lie with their implementation, but rather indicate the presence of more fundamental weaknesses. In fact, only a few countries are considered to have significantly reduced corruption levels in recent decades. The list of achievers varies depending on the observer, but it tends to include Hong Kong, Singapore, and a few other small countries.Footnote ³ The United States, during the late nineteenth and early twentieth centuries, is also an example of a country that transitioned from high to low levels of corruption. However, these changes occurred over long periods of time and were part of broader societal transformations, rather than solely anti-corruption reforms.

The lack of success in anti-corruption efforts and accusations of political motives behind them, as seen in the case of Brazil, should give us pause. The recurrent accusations that anti-corruption efforts pursue political projects that have little to do with their stated objectives, again as it happened in Brazil, where many saw the motivations behind Lava Jato to be less than honest, also suggest that it is time for reckoning, and that we should rethink corruption. More to the point, we should rethink what, at a risk of simplification, I call the prevailing paradigm about corruption. In the decades that followed World War II, sociological functionalism affirmed that corruption may have positive “latent” functions. For example, in certain historical situations, it may have enabled forms of social inclusion when alternatives were absent. A different view gradually took hold toward the end of the 1970s. This developed in the United States using the economists’ methods of analysis and focused on individual behaviors, often observed through the lens of a principal–agent model. Later, theoretical considerations were supplemented by empirical analyses that appeared to confirm that corruption had negative implications and no redeeming qualities, while also shedding light on its presumed determinants and effects. Theoretical considerations and empirical evidence combined to indicate policy solutions to fight corruption. An emerging tool kit of anti-corruption measures gradually became institutionalized, with contributions from a diverse group of actors. There was active involvement from international organizations, such as the World Bank and the United Nations. An influential anti-corruption international nongovernmental organization, Transparency International, was founded in Berlin in 1993. National governments also became involved, and in 1977, the passage in the United States of the Foreign Corrupt Practices Act had important implications and reverberations worldwide. Researchers at universities and other institutions, in the United States and elsewhere, became increasingly interested in corruption, and an industry of specialists on anti-corruption (and, more broadly, on governance issues) developed.

This, using a very broad brush, is what I mean by the current thinking on corruption. How should we correct it? I anticipate the main ideas of this book, together with a brief description of the methodology that I adopt.

Rethinking Corruption as a Contested Concept: Corruption as a Social Construct

I derive most of my conclusions from a detailed analysis, using both primary and secondary sources, of three countries: Brazil, Russia, and the United States. These countries have varied levels of corruption as we measure it, which would rule out the problem of “selection on the dependent variable.”Footnote ⁴ However, the choice of these cases has not been random. As a precondition, I selected countries whose national language I understood at least well enough to be able to read primary sources. These countries were also chosen because I found them instructive in different ways, to the point that observing them over many years has partly shaped my theses. I could then be accused of a selection of cases, if not on the dependent variables, on the covariates, to the extent that, in leading me to the conclusions that I like, they would make me a victim of confirmation bias.

While such accusations could be leveled at all “small ‘n’” studies, it is possible that the three countries that I have chosen represent unique quirks in a world that otherwise manifests regularities of a different type.Footnote ⁵ I aim to convince the reader of the generality of my conclusions in two ways: first, by providing good reasons to believe that they have a degree of generality, in that their motives are plausibly of wide application; and second, by summoning further evidence that supports my theses. I adopt an interpretative framework that is influenced by historical institutionalism and highlights the importance of contextual information. Additionally, my analysis of the country cases is not solely comparative, as I also consider each one of them individually for what they may teach us. Last, while occasionally I will refer to the most recent events or data, I conclude my analysis before the beginning of the COVID-19 pandemic, because of its exceptional nature.

I propose a methodological approach to thinking about corruption that distinguishes between two focuses of analyses. One I call “corruption-as-phenomenon,” and it is familiar as it pertains to investigations that are conditional on a given definition of corruption. Within analyses of corruption-as-phenomenon, we ask questions such as how much corruption there is (e.g., in the form of bribes to public officials) and what its causes and effects are. Most current research on corruption falls under this category.

However, I am mostly interested in a distinct focus of analysis that emerges when we observe the debate on corruption from a distance. There has always been much disagreement about what and who deserves the label of corrupt. This is also evident from the cases that I consider. In Brazil, the anti-corruption protagonists of Lava Jato were eventually seen by many to be corrupt. In Russia, the state used its power to frame the leader of an anti-corruption movement, Alexei Navalny, as a criminal, while it sponsored a friendly anti-corruption movement. In the United States, there has been a concomitant, long-run increase in economic and social inequality. To account for these changes, I consider the concept of “legal corruption,” which goes beyond the currently prevailing narrow definition of corruption. As different as they are, all these cases, which the chapters ahead consider in detail, indicate the contested nature of the concept of corruption.

We should, however, do more than recognize that corruption is an elusive and contested concept, as if such characteristics were ancillary or incidental, and perhaps reflective of a not-yet-achieved perfection in providing a satisfactory definition of what corruption really is. To the contrary, the impossibility of proposing an agreed-upon definition is perhaps what characterizes corruption the best, together with the struggle to influence such an understanding and, at least occasionally, to ensure that its infamous label falls on our enemies and not on ourselves.Footnote ⁶ In this light, corruption emerges as a concept that is always contested and that, as such, cannot be defined univocally. This contested and elusive character of corruption is at the center of my analysis, together with an invitation to go beyond a common understanding of corruption as a set of behaviors, to be appropriately defined, studied, and hopefully eradicated with the right treatment.

Corruption emerges from these considerations as a socially constructed concept, one that may be articulated in different problematic areas. One of them I simply mention without considering it further. It is the problematization of corruption, a term that I use in a loose Foucaldian sense. How and when did it happen that humans started to think about corruption? When did they elaborate such a concept, and when did they start arguing about competing definitions of it? When, in different words, did corruption begin to be considered a problem? I can only speculate that it happened together with the appearance of the first proto-states, which possibly also marked the emergence of forms of codified punishment that could be meted out to persons guilty of such “corruption.” These are just suppositions, as a “deep history” of the concept of corruption has never been written and probably cannot be written due to lack of documentary evidence.

A second element regards the politicization of corruption, that is, the possible instrumentality of accusations of corruption in the political arena (as a “tool of politics,” as I will propose later). The politicization of corruption is part of the broader activity of social construction of the concept, which doesn’t only include intentional uses of corruption as a tool of politics. In fact, many attempts to influence the debate on corruption do not have an explicit political objective. However, and regardless of conscious intentions, they all contribute to our understanding of corruption and have a political effect (this book being no exception). Therefore, as a focus of analysis, the social construction of corruption is general enough to include all the relevant elements (problematization, politicization, and participation in the debate on corruption) that characterize the contested nature of the concept.Footnote ⁷

The thesis that corruption is a socially constructed phenomenon, and that we will never conclusively agree upon what and who is corrupt, may be criticized for being dangerously relativistic. If we deny the possibility of agreeing on a definition of our object of inquiry, perhaps we are left in a vacuum where no interesting statement about corruption might be heard or uttered. I am convinced that this is not the case and that in fact an adequate recognition of the contested nature of corruption is a prerequisite to progress in its study. While much of this book attempts to demonstrate and to give substance to this thesis, here I would like to preliminarily justify it while referring to an important antecedent to the idea that a concept might be subject to endless contestation.

Walter Bryce Gallie proposed that some concepts “are not resolvable by arguments of any kind” and inevitably involve “endless disputes.” An example of such “essentially contested concepts,” as he called them, would be art.Footnote ⁸ Recognizing that art is an essentially contested concept does not stop us from going to museums and, more generally, from participating in the endless discussion about whether what we see there is, or isn’t, an expression of art. We are perfectly able to engage the debate on art at two distinct levels: from a distance, while recognizing that it will never be resolved conclusively; and also from within, where we weigh in with our opinions. If we successfully do so, while distinguishing these two focuses of analysis, our participation in the debate on art may become more mature and nuanced.

Gallie claims in fact that “one very desirable consequence of the required recognition in any proper instance of essential contestedness might therefore be expected to be a marked raising of the level of quality of arguments in the disputes of the contestant parties.”Footnote ⁹ Such conclusion provides the main justification that has been proposed for the introduction of the category of essentially contested concepts, which is perhaps understandable if we consider that philosophers, such as Gallie, are naturally preoccupied with questions that pertain to the quality of intellectual disputes. This line of reasoning certainly would also apply to the debate on corruption. However, I am convinced that, at least in the case of corruption, at stake there is more than the quality of the debate. Corruption is not just any concept. It is, and perhaps has always been in history, the most negative value in politics – on par, in some places and epochs, with crimes of heresy, or of high treason in times of war. Influencing a societal understanding of what and who is corrupt is a very high-stakes game, and we should consider it as such.

To do so, we should keep the two levels of analysis that I propose – that of corruption as a phenomenon, and that of corruption as a social construct – distinct. And since, as I will argue, any definition of corruption as a phenomenon hinges on values and on a normative view of society (or at least, on a standard of action), doing so requires that we as researchers are forthcoming about our values, to the extent that they shape our own understanding of what and who deserves the label of corrupt.Footnote ¹⁰ I advocate for transparency in this regard, recognizing that whenever we discuss corruption-as-phenomenon, we participate in the very high-stakes game of the social construction of corruption. My value judgments, normative views, and perhaps my idea of justice, as they pertain to a discussion of corruption, may be briefly summarized as follows: Democracy is desirable, and a working democracy cannot exist in the presence of pronounced inequalities. The reader will observe these convictions of mine emerge in the pages ahead.

Corruption Is a Tool of Government, and Anti-corruption Is a Tool of Politics

A focus on the social construction of corruption leads to the two main conclusions of this book.

First, corruption is a powerful tool of government, in part because it offers reasons for elites to remain cohesive. These incentives may be in the form of both benefits and punishments. To the first type belong the enticements of corruption, and they are reinforced whenever the ruler has the possibility to assign at will the label “corrupt” and the ensuing punishments. This helps rulers solve an existential problem of control. I use this term in the Madisonian sense that “you must first enable the government to control the governed,”Footnote ¹¹ or, seen from a different angle, that it is necessary for a government to have a degree of control in the choice and in the execution of policies. Control in government is a prerequisite for political order and a vital necessity of any political system. However, corruption as a tool of government has fundamental flaws,Footnote ¹² and a polity should strive to develop better ways of governance. This is easier said than done, and occasionally there may be no concrete alternative solution to the problem of control in government but the use of corruption, as is the case in Russia and Brazil.

My second broad conclusion is that anti-corruption is a tool of politics, because it can be used to pursue a political agenda. Rulers may politicize corruption in a self-serving way, as we observe in the imprisonment for alleged corruption of Aung San Suu Kyi in Myanmar.Footnote ¹³ Additionally, corruption is a tool of politics that is available more widely, because it is a powerful “valence issue,” that is, one of those issues on which most people agree, either negatively (as in the case of corruption) or positively (as it would be for, e.g., “competence”). As such, it is a compelling political rallying cry, as populist movements around the world know well. However, as a tool of politics, anti-corruption has many shortcomings. The case of Brazil suggests that it is rather unpredictable in its outcomes; in Russia, anti-corruption efforts have led to state repression, instead of positive reforms. Perhaps, then, anti-corruption should be avoided as a political platform, but in concrete situations, this might be difficult to do for lack of feasible alternatives.

The two main conclusions of the book – that corruption is a tool of government and anti-corruption is a tool of politics – derive from the distinction between the two levels of analysis discussed earlier. This distinction reveals an interesting symmetry between corruption and anti-corruption, as both are tools, one of government and one of politics. Both have serious shortcomings, but in certain situations, they may be almost inevitable. By acknowledging such symmetries between corruption and anti-corruption, one appears as the obverse of the other.

The Chapters Ahead

This book is divided into three parts. In Part I, I provide a critique of the current prevailing view on corruption and propose a methodological framework to move beyond it. In Part II, I consider the cases of Russia, Brazil, and the United States. In Part III, I take stock of these cases and I discuss and elaborate upon the conclusions of my research.

While arguing that we should think differently of corruption, I have said little until now of the shortcoming of the view that prevails today. Chapter 2 describes its main traits in more details. In particular, the current consensus view on corruption is all but monolithic, and at its “soft edges” we find themes that in fact deserve center stage, such as the elusive and contested nature of the concept.

One aspect of the currently prevailing view is the emphasis on quantifications of corruption, which have been used to research its causes and effects and to gauge the progress of anti-corruption reforms around the world. Chapter 3 is dedicated to these measures. Corruption country scores are an example of so-called Global Performance Indicators (GPIs), which have become popular since the 1990s; they assume that by taking the right initiatives, countries can improve their ranking in a given GPI. However, the available measures of corruption are not well suited to assessing changes of corruption over time. A more general conclusion also emerges from this chapter. In studying social phenomena using quantitative techniques of analysis, it is considered important to draw a sharp line between the definition of a concept, which should come first, and attempts at measuring it, which should be conditional on the definition chosen. However, when measures of social phenomena are successful, they take on a life of their own and contribute to an ossification of the concept they refer to. Consideration of the extent to which the prevailing concept of corruption and its most popular measures have shaped each other also provides a good angle from which to discuss corruption more generally.

Chapter 4 discusses attempts to estimate the effects and causes of corruption. Results obtained using the linear regression models have suggested a list of the determinants, and of the effects, of corruption. They also have implicitly promoted a view according to which we read the observed associations between corruption and other factors in causal terms. I propose a critical assessment of whether the counterfactual of a significantly different level of corruption from the one observed may be legitimate. Corruption emerges from this discussion as a part of a dense historical matrix, one where it is difficult to contemplate changes in individual factors while leaving others unmodified.

Chapter 5 concludes Part I. I first focus on corruption-as-phenomenon and summarize the concept of corruption, contrasting two perspectives that have coexisted. From one perspective, corruption is seen as a degradation of the social body, often within an organic view of society, while in the other, it is defined in terms of public office, for example, as the abuse of entrusted power (or more specifically of public office) for private gain, which is the definition prevailing today.Footnote ¹⁴ Following these considerations, I clarify that all definitions of corruption (as phenomenon) have common ingredients; in particular, they depend on a normative view of the polity. These characteristics may be expressed using a geographic metaphor; later in the book I propose new such “geographies.” I also consider the social construction of corruption and its different actors.

With these premises, Chapter 5 discusses a simple analytical framework that permits us to consider corruption as closely linked to other phenomena and as highly dependent on context. I surmise that historical processes do not proceed linearly but are characterized by historical junctures, which are moments of fast-paced historical change. I distinguish between two types of historical change. One takes place in a “lower sub-system,” which is concerned with day-to-day history. These changes occur without modifications in the ways in which the problem of control of government is solved and political order is guaranteed. The other, which occurs in a “higher sub-system,” implies modifications in that respect. For example, in Brazil, there was much change following Lava Jato, but likely no reduction in corruption. In my interpretation, this is because the need for corruption as a tool of government remained largely unmodified. Lava Jato affected, and even shocked, the lower subsystem, but not the subsystem higher up.

This methodological compass guides me in the second part of the book, dedicated to the description of the cases of Russia, Brazil, and the United States (Chapters 6–8). When considering the first two of these cases, the two important conclusions of the book emerge: Corruption is a tool of government, and anti-corruption is a tool of politics. When considering the United States, I expand beyond a narrow view of corruption and discuss its legal forms.

In Part III, I consider the implications of the case studies. As I stressed, definitions of corruption depend on a normative view of the polity, and such dependence should be recognized when debating corruption. In Chapter 9, I do so, in the context of a discussion of corruption in the United States, arguing for the relevance of new “geographies of corruption,” and, in particular, of legal forms of corruption. My interpretation of legal corruption in the United States is framed within a dynamic relationship between economic and political inequality, which may be mutually reinforcing.

Chapter 10 presents the conclusion that corruption is a tool of government. The cases of Brazil and Russia indicate how corruption can function as a tool of government by providing powerful incentives, both positive and negative, that help maintain elite cohesion and implement policies. These cases demonstrate how corruption can become a trap, as it may be the only viable solution for maintaining control in government, despite its deficiency as a tool. The differences between Russia and Brazil, and other cases that I consider, emphasize the importance of context-dependent information and of case studies.

Chapter 11 argues that anti-corruption is a tool of politics. This is true as soon as we recognize that it is also a tool of government, because assuring the viability of government is an essential political problem. I examine this issue from a wider perspective, viewing corruption as a “valence issue.” Parties, politicians, and civil society organizations may take advantage of the widespread popular opposition to corruption, and campaign on an anti-corruption platform, especially if it is difficult for them to differentiate themselves in other ways. The decreased relevance of ideological differences in recent decades, particularly following the collapse of the Soviet Union, has contributed to the increased interest in corruption that has accompanied the establishment of the current consensus view. Also, for this reason, adopting an anti-corruption political platform has occasionally been an inevitable choice for reform-minded political actors.

I also explore further why anti-corruption reforms have largely failed. Besides Brazil and Russia, I briefly consider other cases, such as Italy, which in the early 1990s witnessed a spectacular anti-corruption campaign, Mani Pulite (“Clean Hands”). These examples highlight the risk of unforeseen consequences of anti-corruption efforts, even when they may be apparently successful. I interpret these difficulties in light of the model of historical change of Chapter 5, which argues that long-term changes require the development of tools of government other than corruption. I also analyze the reforms that took place in the United States over several decades, beginning in the late nineteenth century, which were accompanied by a decrease in corruption.

***

In this book, by emphasizing the elusive and contested character of corruption, I adopt a perspective that aims to be sufficiently detached from its object of study. Anti-corruption efforts struggle against the perverse aptness of corruption in satisfying that prerequisite of government, which is control. Anti-corruption efforts also fall victim to the function it serves as a tool of politics. From this perspective, I derive a pessimistic view on the prospects of anti-corruption efforts that are narrowly defined, on which I present some concluding considerations in Chapter 12. They vouch for a broader approach to the question of corruption, and they touch upon the idea of modernity and one of its key ingredients: the possibility of human agency.