Combating Corruption in the Twenty-First Century: New Approaches

Why do we still need to ask how to combat corruption? After all, there has been no shortage of attention devoted to this issue over the last twenty-five years: academic researchers, policy-makers, international financial organizations, dedicated anticorruption agencies, civil society organizations, investigative journalists, prosecuting authorities, advocacy groups and coalitions, and individual champions have all engaged in the fight against corruption. And they have produced no shortage of strategies and approaches to win that fight: the World Bank has recommended “six strategies to fight corruption,” designed to complement a prior “two-pronged strategy,” in addition to “10 ways to fight corruption”; Transparency International identified “5 key ingredients” to stop corruption; while the World Economic Forum has published “5 ways to beat global corruption,” as well as “3 key steps to end corruption.”¹ The answers seem to keep coming, but the problem remains stubbornly resistant to resolution.

Indeed, we could argue that anticorruption efforts represent a huge policy failure: there seems little evidence that we are much closer to resolving the issue in 2018 than we were in 1996 when James D. Wolfensohn, then president of the World Bank, announced that “we need to deal with the cancer of corruption.”² Moreover, there has been a growing chorus of calls for a fundamental reassessment of how we should understand and combat corruption, often framed in terms of the need to “rethink” existing approaches.³

All this rethinking inevitably begs the question of what the anticorruption movement and its efforts have achieved so far. One somewhat cynical answer is that anticorruption has highlighted a broad consensus that we need to understand better what corruption is, why it occurs, and what we can do to stop it: a sort of intellectual Groundhog Day that keeps bringing us back to the same fundamental questions. As Transparency International's Dieter Zinnbauer has observed,

In other words, we keep engaging in the same kind of circular logic that suggests the best way to reduce corruption is to develop institutional configurations and socioeconomic settings in which public officials act with integrity so that corruption does not prosper.

In seeking to move forward the discussion on corruption and anticorruption, this essay identifies three things we should focus more attention on and three things we should stop doing. It then offers some practical steps that may address some of the shortcomings of current approaches. Given the scale and complexity of the issues under consideration, the essay offers provocations, rather than fully formulated solutions, in the hope that they may not only contribute to the growing calls to “rethink” corruption, but also help reframe the terms of the conversation.

Integrity is often posited as the opposite of corruption, reflected in the widespread use of the term in anticorruption circles: from ngos such as Global Integrity and Integrity Action through Transparency International's National Integrity System (nis) assessments and the Organisation for Economic Co-operation and Development's (oecd) Public Sector Integrity Reviews and Integrity Weeks/Forums, to instruments such as the recently launched Index of Public Integrity. In practice, though, much of the attention devoted to integrity has been implicit: rather than exploring in depth what should be understood by integrity in public life, and how to achieve it, researchers, activists, and policy-makers have often seemed to assume that integrity will result simply from the elimination of corruption.

Predominant anticorruption approaches respond to a logic that does not sit easily with the promotion of integrity. The reason is that policies designed to combat corruption are usually developed as a reaction or response to specific scandals, or else are designed to prevent particular forms of behavior. They are driven by an attempt to address the visible expression of corruption, focusing primarily on institutional configurations or regulatory frameworks, rather than the promotion of prointegrity values among public officials. This means that the practical expression of integrity in anticorruption contexts often reflects this institutional and regulatory focus: for example, Transparency International's nis approach focuses quite narrowly on formal law enforcement as exercised through core institutions (so-called pillars) and corruption-combating agencies. Similarly, the Index of Public Integrity has a strongly institutional tenor, consisting of six components (judicial independence, administrative burden, trade openness, budget transparency, e-citizenship, and freedom of the press) that contribute to the “control of corruption.” The oecd CleanGovBiz Integrity in Practice “toolkit” similarly emphasizes rules and regulation, even when discussing prevention. More promising is the oecd's recent Recommendation on Public Integrity that includes a section on cultivating a culture of integrity – perhaps the single key factor at all organizational levels in building defenses against corrupt activity – but again major emphasis is placed on control, oversight, and enforcement measures.⁵

Yet ensuring that public officials do not behave corruptly offers no guarantee that they will instead act with integrity. It is quite possible to act noncorruptly but also without integrity; for instance, by performing a task with little effort, habitually turning up late to work, or refusing to cover for colleagues. While the absence of corruption does not imply the presence of integrity, it is not so obvious that the reverse holds: if public officials are acting with integrity, they generally cannot – by most common definitions of the term – be acting corruptly. We therefore need a better conceptual understanding of integrity in public life and its relationship to corruption in order to build an effective model of integrity management: that is, the formal framework that ensures that public officials engage in ethical behavior, acting with honesty and fairness while complying with prevailing legal norms.

But the promotion of integrity faces serious challenges, among them the difficulty of defining just what exactly we do mean by the term, compounded by its overlap not only with anticorruption, but also with ethics, morality, and good governance. The oecd refers to public integrity as “the consistent alignment of, and adherence to, shared ethical values, principles and norms for upholding and prioritising the public interest over private interests in the public sector.” However, just as with generic definitions of corruption, such a conceptualization begs a host of questions, not least about the relationship between personal integrity and role-based integrity – as well as between integrity at the individual or at the institutional level – and also the relationship between public and private sectors. Thus, integrity entails complex relationships with other dimensions and can be analyzed from various perspectives.

Drawing on moral and political philosophy, we can identify the core characteristics of personal integrity as: wholeness (thinking beyond just the personal); action that is consistent with principles (doing the right things); morality (doing things for the right reasons); and process (doing things in the right way). Some would add the coda “even when no one is watching” (attributed, apparently in error, to C. S. Lewis) to indicate that genuine integrity does not require any oversight, though such a prospect is wholly unrealistic in real life. Political integrity, meanwhile, encompasses normative justice, openness and transparency, citizen engagement, and impartial authorities.⁶

An effective integrity-management framework, at whatever administrative level, requires mechanisms that reinforce interaction between the personal and political dimensions. However, with few exceptions (such as the work of governance scholar Leo Huberts and associates at the Vrije Universiteit Amsterdam or the emerging oecd agenda) integrity per se is very often effectively little more than a slogan, or else is subsumed without detailed analysis within a broader emphasis on “good government/governance.”⁷

In contrast, Bo Rothstein and political scientist Aiysha Varraich have argued that quality of government understood as impartiality should be seen as the opposite of corruption.⁸ However, impartiality (“when implementing laws and policies, government officials shall not take anything into consideration about the citizen/case that is not beforehand stipulated in the policy or the law”) has little to say about how to address genuine ethical dilemmas or challenges that, for some, represent the only true test of whether individuals act with integrity.⁹ To take a recent example from the United Kingdom, a woman who had spent twenty-seven years married to a British citizen and whose children and grandchildren were born in the country was deported to Singapore after breaching regulations in relation to periods spent out of the country. In a YouGov survey of more than 6,700 uk citizens, 63 percent of respondents felt that the decision was wrong, with just 17 percent believing that deportation was the right decision (even among ukip voters, 50 percent said she should have been allowed to stay).¹⁰ Although the case had seen the uk's policy on residency rights impartially applied, it raises difficult questions about interpretation of the “letter” versus the “spirit” of a law.¹¹ Indeed, “common sense justice,” in psychologist Norman Finkel's sense of reflecting what ordinary people think is just and fair, suggests that in a case such as this, the exercise of impartiality is problematic.¹² While this is just one example, there are countless situations in which specific circumstances or complexities are not prestipulated in a policy or law, leaving public officials having to rely on their individual discretion or interpretation rather than the impartial application of rules. In such cases, integrity – rather than impartiality – is the key virtue.

The second thing we should start doing is pay greater attention to the drivers of individual behavior, in relation to both corruption and integrity. Although behavioral economists have increasingly focused in recent years on experiments that seek to explain corruption, there is still a very widely held assumption among many researchers that, in the words of economists Benjamin Olken and Rohine Pande, “corrupt behavior can be modeled in line with a few general economic principles: corrupt officials respond to monitoring and punishments as one would expect from basic incentive theory.”¹³ Yet in practice, and as behavioral economics suggests, people rarely act as rational cost-benefit optimizers, and their decisions and motivations are subject to a wide variety of biases and influences that are not always coherent or consistent.

Indeed, some psychologists have suggested that fraudulent behavior is often driven not so much by financial incentives as by more complex sets of relationships, including a desire to help or hurt others even when there is no material gain for the individuals involved.¹⁴ “Bounded ethicality” affects how we make ethical judgments: that is, the way that decisions and choices are framed influences our very capacity to see the bigger picture. Experimental evidence suggests that when people assess business decisions, they use different standards and measures than when they consider ethical choices: the former cognitive frame stresses achievement and success as the key decision drivers, making people subconsciously more likely to consider cheating in pursuit of goals. In short, we routinely overestimate our ability to do what is right, and underestimate the extent to which we may behave unethically without meaning to.¹⁵ One recent study – based on four behavioral experiments using a newly developed corruption game – suggests that, contrary to the idea of a slippery slope through which corrupt acts start small and build up over time, some people may find it easier to rationalize leaping off a cliff when a too-good-to-pass-up “golden opportunity” arises than to bear the moral cost of repeated unethical behavior.¹⁶

In general, though, there is a striking lack of detailed work on the individual motivations that underpin corrupt behavior. In contrast to studies of deviance and criminal behavior, social psychologists have with few exceptions devoted little attention to corruption. And yet corruption is manifested in concrete acts that take place in concrete settings, usually (if not always) involving purposive interaction between at least two individuals. If we are to understand better why people decide to engage in corrupt activity, we urgently need to move beyond the reductionist and simplistic idea that it can all be explained by incentives, and that by changing or tweaking those incentives we can address the issue. Linked to this point, we should focus more attention on the role of unwritten and informal social norms as a driving factor behind patterns of corrupt behavior; as highlighted in a recent Chatham House report on collective action and the social norms underpinning corruption in Nigeria: “identifying the specific social drivers of specific collective practices is critical to designing targeted and effective policy interventions to change those practices.”¹⁷

The third thing we should start doing is accept that we can never “win” the battle against corruption in the sense of defeating or eradicating it, and we should therefore set more realistic aspirations for anticorruption interventions. While there have been some widely cited contemporary “successes” (for instance, Singapore and Hong Kong in Asia, Botswana and Rwanda in Africa, Estonia and Georgia in Eastern Europe), such cases represent a strikingly small number in both absolute and proportional terms, and the evidence suggests things may be getting worse elsewhere. As Syed Hussein Alatas long ago observed, corruption “inheres in all social systems…. It affects all classes of society; all state organizations, monarchies and republics; all situations, in war and peace; all age groups; both sexes; and all times, ancient, medieval and modern.”¹⁸ That will remain true, and so the best we are likely to achieve is to “manage” corruption, or to constrain it within more acceptable limits. In particular, we should identify the most egregious and damaging forms of corruption, the ones that cause most social harm at greatest cost, and focus particular attention on measures to combat them. That inevitably means making hard choices about what we should not expend much energy on, rather than sticking to the mantra of “zero tolerance” toward any and all forms of corruption in the pursuit of chimerical aspirations.¹⁹

We should also focus more attention on the feasibility of any reform measures. As economist Mushtaq Khan has argued, conventional anticorruption strategies have sought to improve rule-following across the board, alongside changes to the cost-benefit calculations of public officials, but have generally failed because they pay insufficient attention to the interdependencies and variables that determine what is possible as opposed to what is desirable.²⁰ To be feasible, reforms need to be appropriate to and consistent with the political settlement in question, instead of seeking, for instance, to introduce formal regulatory changes to legal frameworks in situations in which entrenched elites benefit from their control of informal power networks. There is a tendency in much anticorruption programming to equate law with stable government, rather than focusing on the critical question of how to establish the rule of law in those environments where it is lacking.²¹ In line with the social drivers of specific practices, without a better understanding of how the interaction between institutions, beliefs, and behaviors results in a given legal order, it will be difficult to implement successful anticorruption interventions.

From a focus on feasibility, it follows that we also need to pay more attention to variability both in form and outcomes of political settlements, even if we allow for the individualistic version of modernity that underpins the anticorruption vision of “ethical universalism” associated with open-access societies. In practice, even the most well-ordered Western states fall short on some aspects of the ideal type of corruption-free governance, and – more important – they manifest a range of different modalities to achieve similar ends (ranging from constitutional forms of government to territorial organization, electoral terms and systems, accountability frameworks, and so forth). As development scholar Merilee Grindle has persuasively argued, we should be suspicious of the normative connotations associated with “good” governance and its ever-growing list of requirements, and focus instead on the practical organization of actual governance tasks.²² That means accepting messiness and ambiguity, reflecting the complexity of modern political organization, as well as the need to understand better the interplay between microlevel interventions and macrolevel drivers of historical development.

In order to strengthen our assault on corruption, there are three things we should stop doing, each representing a logical extension of the discussion to this point. First, we should stop talking about corruption as if its meaning were self-evident. For a variety of reasons, the way the term has come to be used by academics and practitioners, as well as by journalists and political commentators, acts as an obstacle to moving the anticorruption agenda forward.²³ A main drawback to discussing “corruption,” without any adjectives, is that we cannot reach any kind of consensus, other than at an abstract or generic level, over what it comprises. To be sure, there is increasing reference to some variation on the now well-worn formulation that corruption is the “abuse of public office for private gain”; but this definition suffers from two principal problems. First, since what should count as “abuse” is not itself defined, the definition effectively begs the question: what constitutes the alleged transgression of what norm, and who decides? Second, even if we could agree on what we mean by abuse, the definition still encompasses such a vast array of different kinds of activity by different actors in different settings that it is not helpful in any operational or policy-informing sense.

Definitions in other fields often provide what might be termed “core” or “umbrella” terms, which are then taxonomically subdivided. For instance, both condors and wrens are birds, but ornithologists have no difficulty distinguishing between them. Alternatively, to use the analogy most frequently applied to corruption, cancer always describes abnormal and uncontrolled cell growth, but encompasses over a hundred different diseases.²⁴ When it comes to corruption, though, we seem somehow stuck at the generic level: much of both the research and the associated anticorruption strategies developed over the last twenty-five years have signally failed to engage in the critical differentiation of pathological characteristics that distinguish some forms of corruption from others. If corruption is a form of cancer (or some other disease), then corruption oncologists need a more sophisticated understanding of its dna if they are to develop effective responses; but discussions of corruption and how to combat it often proceed as if such efforts are a tiresome or annoying distraction. It is as if many who seek to combat corruption set out from the proposition that we sort of know what we mean, so let's not get too hung up on the definitional niceties; in the words of one scholar:

While many scholars and anticorruption practitioners do indeed mean “that cluster of stuff” when talking of corruption, others see the term as encompassing different or additional dimensions (for instance, legal uses of power that nonetheless betray the “democratic transcript” by violating the rationale and spirit of public rules).²⁶ Ultimately, the likely impossibility of establishing an uncontested, yet rigorous, definition means that corruption, unless we specify what precisely we understand by it, ceases to have any clear referent in research or practical policy terms. However, in reality, we rarely see such specification;²⁷ instead, at best, there has been a tendency to develop dichotomous distinctions (grand/petty, political/administrative, systemic/sporadic, individual/institutional, extortive/transactive, need/greed). Not only are modalities of corruption more complex and flexible than suggested by such binary schema, but there tends to be a separation between, on the one hand, work that seeks to identify “types” of corruption and, on the other, policy-oriented approaches to combat corruption writ large.

But perhaps just as significant, this lack of clarity means that the term can easily be pressed into political use as a descriptor of whatever is unpopular. We begin to see corruption everywhere and in everything, the root cause of any form of failure in any political setting: popular protests against the alleged corruption of political leaders have become common throughout the world, regardless of regime type. In turn, mutual accusations of corruption have become the stock-in-trade of political contestation, as was starkly evident in the 2016 U.S. elections, and as has long been the case in post-Communist regimes and elsewhere. As a result, the term has become ever more devalued even as it is ever more widely used, ultimately serving as little more than a boo word.

Our failure to specify what exactly we understand by corruption is reflected in a tendency to amalgamate all forms of corruption into a value that can then be measured within a jurisdiction. This brings us to the second practice to stop: the near exclusive focus on nation-states as our unit of analysis; and with it, the increasing production of rank indices measuring the amount of corruption in any one state compared with another. There is now a very extensive literature on the problems of measuring corruption, but relatively little of it questions the utility or relevance of focusing on nation-states. While there are a few measures that look at the subnational level, notably from the Quality of Government Institute at the University of Gothenburg, the vast majority of approaches both to measurement and remedies are pitched at the level of individual countries. In the words of political scientists Alexander Cooley and Jason Sharman:

In many ways, such a focus on nation-states is entirely understandable. States are so well established as units of political analysis, given their dual claim to sovereignty and legitimate authority, that it is natural to focus on individual state corruption and state efforts to combat it, just as we look at and compare a host of other governance indicators at the state level. However, there are clear problems with such an approach. To begin, it rarely works in practice, since most national-level measures of corruption provide only a single aggregate indicator. Moreover, it can still mask potentially significant variation within countries; as political scientist Staffan Andersson has observed, “actual instances of corruption may vary spatially – both subnationally and across government levels and sectors – a factor not detected by studies relying on single-country scores.”²⁹ The obvious case is Italy, with its stark differences between North and South, though these variations are also evident both in very large jurisdictions like Russia and India and in small countries like Sweden and the United Kingdom.³⁰ Research shows that 80 percent of the variance in the outcomes of aid projects occurs within countries rather than across them, further undermining the idea that there is anything approaching a uniform “level” of corruption within nation-states.³¹

But most important, the focus on nation-states fails to reflect the reality of how some of the most egregious forms of corruption operate in practice, particularly those that entail shifting value from one jurisdiction to another. We know that kleptocracy most often relies on the collusion of rich countries: “the bigger the financial center, the more dirty money flows through it.”³² While it is true that effective action against such activities would require sovereign nation-states to come together in a concerted manner, that is unlikely to happen if we systematically underplay the interconnectedness of how some forms of corruption function and the facilitating role of so-called clean countries. As Cooley and Sharman have observed, the implication for how we research and combat corruption is that we need “a shift in the unit of analysis, to transnational networks, rather than just states,” a point underlined by the revelations in the “Panama Papers” and “Paradise Papers.”³³

Third, we need to stop searching for unicorns. By this, I mean any attempt to identify “the answer” to how we should combat corruption, exemplified by the various lists mentioned at the start of this essay. There is now a consensus that one-size-fits-all approaches are doomed to failure, and yet the temptation to develop prescriptive approaches remains deeply embedded and anticorruption “toolkits” abound. Even if those once offered by the United Nations, for example, are no longer peddled quite so hard, the language of anticorruption tools and toolkits is still widely used, as is the invocation to establish anticorruption agencies with common standards.³⁴ It is essential, of course, to outline some core standards and expectations against which to judge progress, but research increasingly emphasizes the need to pay due attention to specific contexts and how these shape the implementation of any given reform measures. Equally, the “paths to success” in those countries deemed to have done well in combating corruption can vary significantly, calling into question any notion of one specific or “correct” route: as the eighteenth-century biologist and Catholic priest John Needham observed, there are more ways to heaven than one. The evidence suggests that, historically, crisis and existential threats, gradualist reform, visionary leadership, and/or popular demands can all be key factors – depending on their interaction with other circumstances – in moving from particularism to the more universalistic and impartial provision of public goods that characterizes low-corruption jurisdictions.³⁵

There is, then, no single route that needs to be followed, just as apparently similar circumstances in different countries can result in very different outcomes. Rothstein convincingly argues against precisely the institutional toolkit approach that seeks a “magic key” to unlock incremental anticorruption measures.³⁶ He uses the examples of Sweden and Denmark to illustrate how a “big-bang” approach to major administrative reorganization in the mid-nineteenth century – following crushing military defeats that threatened their very survival – better accounts for the move from “limited” to “open-access” orders. However, in Spain, which also suffered a catastrophic military defeat at the end of the nineteenth century, the response was very different: in this case, rather than a move toward more impersonal, universalistic forms of administrative organization, the crisis brought forth a so-called iron surgeon in the form of Miguel Primo de Rivera, establishing a model of Spanish dictatorship that would persist for much of the twentieth century. Ultimately, and as unsatisfactory as it may seem, it is hard to escape the conclusion that a very significant element of contingency plays into how different administrative orders have come about. For all that comparative historical and institutional sociologists – building on Barrington Moore's classic Social Origins of Dictatorship and Democracy – have persuasively identified different routes to the modern world, we still struggle to unpack the precise mechanisms that explain the relationship between long-term structural factors and short-term agency.³⁷

Given these arguments, the challenge is to identify what kinds of specific initiatives may work in combating corruption. An essential first step is to specify both the type of corruption in question and its locus. That means paying far more attention to the modalities of different forms of corruption and how they change and mutate in response to wider sociopolitical and economic developments. Different strategies will be required to address different types of corruption. If we wish to concentrate on tackling transnational corruption, for instance, then we need a more sophisticated understanding of how licit and illicit practices interact to establish global networks populated by individuals who exploit channels made available through the operations of banks, shell companies, realtors, and so forth.

Of particular note, the emergence of what has been termed the “post-modern state” has led to a blurring of conventional divisions between different spheres of activity, notably the public and private sectors. Despite continued contestation over the nature and meaning of processes associated with both “globalization” and the “hollowing out” of the state, the fact is that in many states, there no longer exists a clear separation between the respective remits of public and private providers: not just in terms of policy delivery, but increasingly in terms of policy design, especially in relation to financial and regulatory matters.

This point matters greatly for any attempt to tackle transnational corruption because corrupt individuals have been able to adapt and change in response to these broader developments, exploiting new opportunities in a continuing effort to outsmart the regulatory measures designed to curb their activities. Notably, changes in the global financial architecture associated with the way that money can be moved internationally have not only made it much harder for individual states to enforce effective regulations, but have also created new opportunities for illicit international networks to hide the proceeds of corrupt activities. At the same time, the increased interpenetration between public and private sectors, associated with the rise of what have been variously described as “business politicians,” “flexians” (an elite professional class who serve multiple, overlapping roles ranging across government and private enterprise), or “globalized individuals,” undermines traditional forms of accountability.³⁸ In this emerging and fluid environment, the ability of individuals to move between a range of roles and nationalities, coupled with quantum leaps in information technology, offers new opportunities for malfeasance.

The implication of such developments is that successful efforts to tackle the increasingly transnational nature of much contemporary corruption, often structured through intermediaries operating in the legal and professional world, requires much greater international cooperation and shared activity than has been the convention in anticorruption programming. A transnational strategy must put greater emphasis on the policy role of such international bodies as the Financial Action Task Force, supported by its equivalent regional groups,³⁹ as well as more extensive use of such instruments as beneficial ownership regulations, unexplained wealth orders, and the United Nations Office on Drugs and Crime's International Money Laundering Information Network. Twenty eu member states have recently announced an agreement to set up in Luxembourg the European Public Prosecutor's Office, designed to tackle cross-border fraud, and some, including Judge Mark Wolf in this volume, have called for the establishment of an International Anti-Corruption Court, modeled on the International Criminal Court.⁴⁰ Clearly, the role of such instruments in turn raises questions about accountability, transparency, and civic participation, but it is clear that without more concerted cooperation across states, nationally focused responses are unlikely to succeed in managing corruption.

A second observation about practical interventions again relates to scale, but this time it focuses on the subnational rather than transnational level. It is noteworthy that two of the most widely cited examples of positive anticorruption measures, Hong Kong and Singapore, are effectively city-states. Other “success stories” have also involved various agencies working together to clean up cities, as in Lviv, La Paz, Monterrey, and Ciudad Juárez.⁴¹ That points to the need to focus more attention on anticorruption initiatives in urban centers. Just 3 percent of the earth's landmass is urbanized, yet cities are responsible for around 70 percent (and growing) of overall primary energy consumption, reflecting their status as the prime drivers of the global economy: 50 percent of global gdp is generated by 380 cities in the developed world (20 percent from 190 cities in North America alone), and just 20 cities are home to one-third of all large corporations and nearly half their income.⁴² Dieter Zinnbauer is one of the few to have noted that cities matter in fighting corruption, pointing out a “double blind-spot: anti-corruption analysts and advocates do not pay sufficient attention to cities, and urban practitioners do not pay enough attention to corruption.”⁴³ Writing for The Guardian, Jack Shenker observes that

In June 2017, the Center for Advancement in Public Integrity at Columbia Law School hosted the Global Cities 11 conference, which focused on the increasing use of data analytics to combat corruption.⁴⁵ The conference brought together anticorruption leaders from Bogotá, Cape Town, London, Melbourne, Miami, Montréal, New York, Paris, Rio de Janeiro, and San Francisco to explore how data driven approaches in areas highly vulnerable to fraud (benefits, human resources, procurement, campaign finance) can help identify outliers that may indicate untoward activity. Anticorruption analysts have put great hopes in the use of open data to help fight corruption, notably through the auspices of the Open Data Charter, following recognition by the g20 Anti-Corruption Working Group of its potential to follow financial flows and highlight irregularities in public contracting and procurement processes in particular.⁴⁶ The use of open data, of course, is not a magic bullet, and the extent to which it offers a route forward will be conditioned by other factors (including the capacity to interpret and act upon it). Whereas data analytics are clearly of significant potential relevance in “world cities” where illicit money flows enable corrupt networks to operate, they are less likely to be helpful in war-torn centers where law and order has effectively collapsed, or where bribery for access to services is routinized and there are strong links between corruption and generalized violence and organized crime.

The key point, however, is that successful anticorruption initiatives can start small: there is no ineluctable need for them to be led from above by national authorities. That is encouraging, since it suggests that there can be a possibility of progress even when political will at the national level is lacking. Increasingly, anticorruption researchers and programmers are recognizing that the need to consider context means that initiatives need to be highly targeted, both in terms of scale and sectoral focus.

In addition to looking in more detail at the prevalence and operation of corruption in cities and at other subnational levels, there is a growing realization that we need to pay greater attention to sectoral differentiation, exploring the pathologies of corruption risk in specific fields. This line is further developed, with notable advances in studying procurement processes, as well as initiatives such as the Transparency International global programs on defense and security, pharmaceuticals and health care, and mining, among others, together with a growing recognition of corruption in fields like education and professional sports. However, in general anti-corruption programming, there is still too much emphasis on the public sector versus the private sector, coupled with calls for generic, national anticorruption plans and strategies.

The argument I have sought to make is that if we want to make progress in tackling corruption, we need to make fundamental changes to our approach. Most important, we should stop talking in global, generic terms about corruption and be much more attentive to which precise issues we are concerned about, where they occur, and how they operate. That means taking proper account of how changes in both the international architecture of global trade and finance, and also the design and organization of modern states, affect the nature of and possibilities for emerging corrupt exchanges. We therefore need to rein in our ambition and be more realistic about what it is possible to achieve and at what level. We need to focus more attention on appropriate units of analysis, both in terms of research and in policy-formulation; that means better understanding when and why individuals engage in corrupt activities, how their actions are shaped by social norms, and how those norms can be changed. Equally, we should pay much more attention to what we mean when we talk about integrity, recognizing that it does not result simply from removing corruption. In short, we need to change the terms of the conversation and accentuate the positive, rather than only trying to eliminate the negative.