Getting to Accountability: A Framework for Planning & Implementing Anticorruption Strategies

Corruption is a complex problem with enormous political salience. It is therefore not surprising that the solutions academics proffer for addressing corruption – long-term structural remedies that may not mature for decades or quick solutions that are almost certain to founder as they are battered against preexisting political conditions – frequently leave policy-makers dissatisfied.

The first approach to combating corruption begins from the premise that it has deep and structural roots in culture, social inequality, and the (un)rule of law.¹ Effectively targeting corruption when structure is the driver requires a “big bang,” a critical juncture, or a historical turning point momentous enough to pull a country off its current path. Describing the Korean, Japanese, and Finnish cases, political scientists Eric Uslaner and Bo Rothstein have suggested that external stimuli from the Japanese occupation, American postwar occupation, and the Soviet threat led all three countries to invest heavily in education as a means of nation-building. Given the strong ties between education, systemic inequality, trust, and corruption, this investment paid big dividends for development and, incidentally, for anticorruption.² But it took catastrophic war or the threat thereof to jolt societies into action.

Other scholars in the structural-change school suggest that elite displacement may be the key causal mechanism to combat corruption. A generational shift that changes policy priorities might do the trick: universal social-welfare policies, for example, have in recent years helped generate a change in the corruption equation in the new democracies of the developing world, whether by weakening old patronage practices or reducing inequality and thus enhancing trust.³ Scandal, economic shock, or war may lead to elite replacement. The trouble with these structural theories of change, of course, is that anticorruption gains are often purely incidental, external sources of change cannot be conjured from thin air, and triggers for change (such as wars, genocides, or regime changes) may be even more damaging than the underlying disease of corruption. Further, the time horizon for these structural improvements is usually at least several decades long – hardly the stuff for today's results-oriented reformers.

At the other end of the spectrum, a second group of anticorruption advocates offers up immediate remedies for symptoms of corruption. But this literature all too frequently suggests specific tactics without explaining how those solutions will work to fight corruption. Jeremy Pope's influential Transparency International handbook on national integrity systems, for example, suggests the establishment of “integrity pillars” – institutions needed to fight corruption – but does not provide a theory of what each is intended to accomplish or how to prioritize among them.⁴ One reason for this lack of clarity may be that the prescriptions appear self-evident: efficient courts or independent auditors seem preferable to their dysfunctional or subservient alternatives, after all. But little is in fact known about what actually drives change in corruption levels in the short term, or how these solutions build on each other. The result is a laundry list of one-size-fits-all remedies, provided without much guidance for implementation, sequencing, or concern for the systemic whole, which at best will correct topical maladies.⁵ Even broader national anticorruption strategies, very much in vogue these days, are frequently developed without great thoughtfulness about the prescribed reform measures and the changes that they are designed to generate.⁶ There is no theory of change undergirding their implementation, meaning that best practices may be plopped down without much consideration of local conditions, leaving them vulnerable to co-optation by local power structures.

This essay uses anticorruption success stories to argue that strategic, incremental, and iterative accountability reforms offer a pragmatic alternative to deeply structural or highly specific institutional anticorruption approaches. The first section describes the relationship between policy bursts and anticorruption equilibria, providing historical experiences from countries where small bursts of anticorruption efforts accumulated into lasting shifts in the local accountability equilibrium. The second section draws on these cases to argue that we already know a great deal about the accountability systems required to generate lasting shifts in the corruption equilibrium. Such knowledge may help us to speed up and focus contemporary anticorruption efforts more strategically. The final section uses contemporary case studies to propose a strategy for tackling the bottlenecks to effective accountability: the goal is to pursue reforms that alleviate the most binding constraints (“bottlenecks”), and thus produce the “biggest bang for the reform buck,” and to do so iteratively, so that one success builds on another.⁷

Cold realism is needed. Many of the countries that today have reasonably effective anticorruption systems stumbled across them by historical happenstance. Historical cases of a significant and lasting shift in corruption levels are few and far between. Those we do find appear to have arisen through incrementally implemented changes in effective governance rather than through wholesale systemic reforms targeted specifically at corruption.⁸ Depressingly, the shift from “closed-access” political systems marked by particularism and distrust to open-access systems with inclusive political institutions, universalism, and formalized trust is rare; economist Douglass North and colleagues note that only about twenty-five countries have made that leap in the past two centuries. Indeed, closed-access systems are the historical norm.⁹

But realism need not imply nihilism. Building on lessons from past successes, it should be possible to introduce the kinds of improvements in transparency, oversight, and sanctioning power that cumulatively add up to a shift in the overall corruption equilibrium. The shift is likely to be a multigenerational effort, with the possibility of reversals along the way. But even if the process of strategically developing accountability institutions does not guarantee movement all the way from a closed to an open-access order, such a process may nonetheless be able to move countries to a normatively preferable new equilibrium. And the accumulation of knowledge from past experiences should enable us to formulate an anticorruption strategy that permits quicker progress on anticorruption today than was possible in the past.

Before proceeding, it will be useful to define policy bursts and equilibrium shifts. Policy bursts are policy interventions that may have immediate effect, but whose effects can peter out if left untended or unused. For example, passage of a freedom of information law may inspire a short-lived moment of transparency before bureaucrats learn tricks for blocking inconvenient inquiries. New prosecutorial tools (such as anti – money laundering laws) may spawn new cases until criminal defense lawyers adapt, prosecutions hit new roadblocks further along in the judicial process, or criminals find new ways to transfer ill-gotten gains. Anticorruption agencies (acas) may achieve initial results but then founder, as they have in most countries, because of their insertion into a hostile environment marked by weak political will, low investment, and internal corruption.¹⁰

This is not to say that only isolated policies are at risk of petering out. Wholesale across-the-board approaches in which a broad range of instruments are simultaneously adopted can also prove to be short-lived. The Chinese case is emblematic: General Secretary Xi Jinping's massive anticorruption campaign introduced a variety of compliance-based tools and even changed some officials’ behavior. Yet because the campaign has privileged sanction over prevention and compliance over a broader norm of integrity, its anticorruption efforts have naturally been applied selectively, with political motivations.¹¹ In the political-economic context of a Communist party-state, prevention has few political payoffs, while enforcement has more. The end result is that Xi's anticorruption crusade has not led to a lasting change in behavior and is unlikely to move corruption to a stable new equilibrium. Similarly, a broad “big-bang” anticorruption campaign in Mexico under President Vicente Fox failed to develop into an equilibrium shift before his leadership and his party were discredited by corruption allegations of their own.¹²

Equilibrium shifts are transformations into new self-sustaining, societal states of mind. It takes a big push to move out of one equilibrium into another, because equilibria are sustained by norms, routines, and institutional patterns of behavior. In the most negative equilibria, institutions are subverted by the very governance flaws they ostensibly should combat, and status and connections are the core determinants of public-goods provision. A lack of trust becomes the guiding norm governing behavior across the society, and corruption is only one expression of it.¹³ The most positive accountability equilibrium, by contrast, is a societal state marked by ethical universalism; reciprocal accountability; and fairness and public integrity.¹⁴ A positive equilibrium has been described as a virtuous cycle in which empowerment encourages citizens to participate, institutions function, citizens feel empowered to use them, and citizen participation in turn strengthens institutions.¹⁵ In a word, Denmark.¹⁶

The goal is therefore to spur a shift from a “corrupt equilibrium in which it can be irrational and even dangerous to be honest”¹⁷ to an equilibrium in which accountability “feeds on itself” in a virtuous cycle.¹⁸ There will be intermediate equilibria along the way; the goal is to move stepwise from one equilibrium to the next, addressing some of the constraints to successful anticorruption in a strategic manner aligned with local conditions.

When things go well, policy bursts will collectively add up to an equilibrium shift. Perhaps the best known example is the United States, where during the Progressive Era (from the late 1800s to early 1900s) a largely uncoordinated and unplanned series of accountability efforts took place.¹⁹ Driven by a vague (and sometimes ugly) malaise, public pressures led to an accumulation of unrelated incremental institutional reforms, such as regulation of the trusts, elimination of patronage hiring in the civil service, restrictions on corporate campaign contributions, and an end to boss-driven politics. Although many of these changes began in the late nineteenth century, they only precipitated a significant shift in the accountability equilibrium between the 1920s and the New Deal.²⁰ Summarizing a complex history, economists Edward Glaeser and Claudia Goldin use quantitative measures of press coverage of corruption to demonstrate an arc-like pattern: corruption rose steadily from 1815 to 1850, but began falling after 1870, reaching a stable lower-corruption equilibrium by the 1930s, where it remained until the 1970s (when the authors ceased data collection).²¹

Sweden and much of Scandinavia underwent similar incremental, decades-long equilibrium shifts. Their experience shares with the United States’ a lengthy transition period, little direct policy focus on corruption per se, and broad reforms across a wide range of institutions that altered the calculus in favor of collective action. Bo Rothstein, for example, has described how over a roughly four-decade period in the mid-nineteenth century, Sweden, reacting to crushing military defeats, undertook multiple dramatic institutional changes toward establishing good governance: strengthening the civil service, removing patronage appointments, enhancing oversight of political leaders, and ensuring genuine political contestation. These were “indirect” reforms in that few of the new policies or institutions directly targeted corruption. But all of the changes were driven by elite concerns about the importance of genuine institutional change.²² As in the United States, the incremental reform process added up to a significant shift in the corruption equilibrium, as cumulative incremental policy bursts accumulated into a lasting equilibrium shift.

Bursts of anticorruption policies do not always result in an equilibrium shift, of course. For all of the heroic stories of anti-corruption efforts carried forward by daring and pioneering muckrakers, prosecutors, and police, there are at least as many cases in which apparent successes proved superficial, tenuous, and reversible.

The Italian case is illustrative. A number of factors made Mani Pulite (“Clean Hands,” an anticorruption investigation that began in 1992) function better and more effectively than past efforts, leading to the investigation of more than six thousand individuals, including more than five hundred members of parliament and five former prime ministers.²³ Greater internal independence among a new generation of investigating magistrates (giudici ragazzini); the popularity of their antimafia efforts; the end of the Cold War and the concomitant decline in the relative importance of political stability; and strong new guarantees of judges’ independence all contributed to making Mani Pulite possible.²⁴ Public support for the prosecuting magistrates spurred investigation, especially after parliament banded together early on to protect itself from prosecution.

Yet Mani Pulite's trajectory became more sobering after this initial burst of activity. Investigating magistrates lost the confidence of the public, in many cases because they were seen as overzealous and politically motivated. Many politicians and businessmen implicated in the case – not least Prime Minister Silvio Berlusconi – thrived in the obliterated political landscape.²⁵ Ineffectual prosecutorial instruments let hundreds of defendants escape under the statute of limitations, leading criminals to have a “sense of impunity.”²⁶ The legal changes that might have followed such a massive investigation were blocked, and Berlusconi's government sought to reduce the allegedly arbitrary powers of judges.²⁷ The constant media bombardment of corruption stories may actually have caused a “saturation effect,” increasing public tolerance for corruption.²⁸ Corruption became even more sophisticated, as political scientist Alberto Vannucci has written: “Corrupt politicians, public servants and entrepreneurs have learnt the lesson … developing more sophisticated skills and [techniques] to practice corruption with higher probability of impunity.”²⁹

Corrupt elites worked proactively to weaken accountability, taking advantage of the fact that the policy burst was relatively isolated in a single institution. Especially once public attention turned away from the investigations, politicians did all they could to make prosecutors’ jobs more difficult: strengthening evidentiary protections, decriminalizing accounting fraud, reintroducing parliamentary immunity, eliminating sentencing rules, and reducing statutes of limitations in corruption cases by more than half.³⁰ The end result was that this anticorruption burst did not yield an equilibrium shift, and there was a modest decline in perceptions of corruption control in the two decades that followed Mani Pulite.³¹

Uncertainty about whether policy bursts will accumulate into a new and improved equilibrium is also evident in contemporary Brazil. In the wake of the 1985 transition to democracy and the drafting of the 1988 constitution, Brazil has been slowly undertaking improvements in its accountability framework in response to a combination of scandals, bureaucratic innovations, and democratic pressures. These improvements included changes to campaign finance and congressional spending rules that emerged out of scandals like President Fernando Collor de Mello's 1992 impeachment and a massive 1993 congressional budget scandal. They included a shift away from clientelistic delivery of social spending through adoption of the Bolsa Família conditional cash transfer program. Improvements also arose from responses to unrelated policy challenges, such as hyperinflation, the remedies for which included greater oversight of tax revenues, better controls over spending, and enhanced regulation of financial institutions. International pressures also contributed to enhancing institutional effectiveness, including through the implementation of strong anti – money laundering laws and the creation of specialized courts for financial crimes.

One consequence of these gradually accumulating institutional improvements has been a number of high-profile investigations of corruption since the turn of the century. These gains have not yet been met, however, by equally strong judicial performance, suggesting that the next bottleneck on the way to a new accountability equilibrium is the court system's chronic ineffectiveness. Meanwhile, strong pushback has arisen from antireform groups, notably politicians who benefit from the status quo and whose staying power has been enhanced by the impunity guaranteed by the judicial system. Although there are hopeful green shoots of a new equilibrium, it is unclear whether Brazil's incremental gains are the sign of an impending shift or whether the country will instead come to resemble post – Mani Pulite Italy. One thing is clear: achieving a more positive new equilibrium will require strategic, conscientious, and informed efforts beyond the current headline-grabbing Lava Jato investigation.³²

As we think about how best to engender a corruption equilibrium shift, it may be useful to focus on accountability rather than on corruption per se. One important reason is that focusing on accountability over corruption may expand the constituencies of political support : after all, accountability has many possible beneficial outcomes beyond anticorruption alone, including policy efficiency and effectiveness, which may be desired even by corrupt incumbents. Accountability is the right to hold other actors to a set of standards, judge whether those standards were met, and impose sanctions if they are not.³³ Accountability generates the desired performance through answerability (those governing are obliged to respond continuously for their acts and omissions) and enforcement (the imposition of sanctions for failing to meet public standards).³⁴

The fact that various mechanisms and bodies can impose accountability makes accountability both polysemic and multidirectional. It is polysemic because it can be imposed in an almost infinite combination of ways: political (removal of ministers, elections); social (egg-throwing, public shaming, reputational costs); legal and criminal (fines, jail time); or bureaucratic (limiting salary, promotion, or tenure or demanding additional information and paperwork). Accountability refers to an actor's assumed responsibilities, which could be formal (the actor pledges not to violate the constitution) or informal (the incumbent should not abuse the stature of the office of the presidency); just as the sanctions imposed may be both formal (a bureaucrat is subpoenaed by congress) and informal (the bureaucrat is not subpoenaed, but knows that there is a long tradition of legislative oversight). Effective accountability typically involves some mixture of judgment (politicians’ performance as judged by voters), norm (what is corrupt behavior), and law (officials’ performance must accord with their legal responsibilities).

Accountability is multidirectional rather than horizontal or vertical: seldom does accountability come about solely from a single agency blowing the whistle and punishing another horizontally, or from voters responding to politicians vertically. More likely, some combination of whistleblowers or media reports triggers accountability processes, which are kept in motion by simultaneous pressures across agencies and from society. It is also multidirectional, because effective accountability does not only entail sanctioning improper behavior but also monitoring and investigating government conduct; and accountability does not solely mean punishing transgression, but avoiding possible transgressions ahead of time and following up if they occur anyway. The institutions and mechanisms that might potentially impose accountability include everything from the most predictable – police, prosecutors, judges, anticorruption agencies, and accounting tribunals – to those not included in conventional thinking about accountability, such as securities regulators, central banks, and revenue authorities. The relevant bodies will also vary by country.

Yet whatever the idiosyncrasies of individual country-level experiences of accountability, going back to first principles suggests that accountability (a) is the outcome of transparency (t), oversight (o), and sanction (s), all of which are moderated by the degree of institutional effectiveness (e), tempered by the degree of political dominance (d). The equation

is widely applicable and can be used at various levels of analysis, from county boards to national legislatures and across or within different policy sectors. The equation allows practitioners wide latitude in determining how to achieve the objective of accountability, but nonetheless provides a structure that can guide strategic policy choices.

The first component is transparency (t), defined in its most essential sense as public access to government meetings, procedures, and information. Transparency gives public agencies, private individuals, and nongovernmental organizations the information they need to evaluate the government's performance on whatever criteria those groups find most relevant. Obviously, transparent data can be made more useful to citizens in any number of ways: At its worst, public disclosure can sometimes be no more than an elaborate ruse manipulated by bureaucrats. At its best, however, transparency assumes an inclination toward information-sharing, easily accessible and timely provision of data, and the ability to verify data across sources, meaning that information is both visible and usable for drawing inferences about government actions.³⁵

The second component is oversight (o), meaning that government functions are subject to surveillance that gives public or private agents the right to evaluate a government's performance more intensively than by simply accessing data furnished by the government itself. Ideally, oversight would be almost unlimited, and all government accounts, processes, and agents would be susceptible to random or targeted audits. Oversight is likely to be most effective when it relies on the reinforcing perspectives provided by multiple overlapping accountability bodies operating independently but conjointly.³⁶ Ideally, this web of accountability agencies would be able to operate in “fire alarm” and “police patrol” modes simultaneously, reacting to unexpected revelations but also continuously probing vulnerabilities so as to unmask inadequate or inappropriate performance.

The third component is sanction (s). Cesare Beccaria, Jeremy Bentham, and economist Gary S. Becker have all suggested that the costs of committing criminal acts factor heavily in individuals’ decisions to engage in wrongdoing. But sanctions also serve a societal role: effective ones may ultimately be less about altering the individual calculus of whether to commit wrongdoings than about generating societal trust.³⁷ Because mutual trust is so important to anticorruption, more important than punishment for a single individual is: 1) demonstrating that there is a societal norm at work and restoring it to its proper place; and 2) the iterative process by which transparency, oversight, and sanctions together point to underlying dynamics that contribute to governance failures and provide clues toward how best to realign institutions and incentives to deter such abuses.

Transparency, oversight, and sanctions are moderated by institutional effectiveness (e) and political dominance (d). Three factors play a central role in institutional effectiveness: 1) state capacity (a professional bureaucracy with the ability to implement policy without undue external influence);³⁸ 2) a robust institutional toolkit, including relevant laws, mutually supportive bureaucracies, and adequate budgets; and 3) citizen engagement, which is a force multiplier for transparency, oversight, and sanction. Political dominance (d), on the other hand, plays a negative role by diminishing the incentives for active oversight or energetic sanction. All other things equal, the more agencies are dominated by the incumbent party or government allies, the less likely they will be able to fulfill their accountability function.³⁹ Political dominance is often associated with lack of political will to combat corruption; reciprocally, increased political competition often brings with it the will to enforce the law energetically and address its shortcomings when necessary.

A frustrating finding from years of research is that anticorruption programs actually work best where corruption is lowest.⁴⁰ When society is stuck in a low-level equilibrium in which corruption benefits the corrupt and low levels of interpersonal trust give the noncorrupt little reason to work collectively to curb corruption, it can be very hard to overcome dominant political interests and impose effective accountability.⁴¹ But one of the most significant implications of the accountability equation is that progress may be nonlinear: as institutional effectiveness improves or political domination declines, small gains in transparency, oversight, or sanction may redound to outsized outcomes. The result is a classic story of punctuated equilibrium: institutional improvements usually accumulate in a slow and incremental manner, but shifts in accountability equilibria may occur with astonishing rapidity.

The World Bank's Control of Corruption indicator ranks only thirty-five countries – roughly one in six nations in the world – in the category of Denmark.⁴² How might policy bursts be designed to push more countries into this high-performance equilibrium, or at least closer to it? Further, is this goal achievable in our lifetimes?

The challenge is daunting. The topranked countries have somehow managed to merge two powerful and contradictory impulses: maintaining a strong and capable state but constraining it via law and democratic choice.⁴³ The twenty countries at the top of the World Bank's Control of Corruption indicator are all rich, mostly small, and predominantly European. In the past twenty years, only four countries have joined the top-twenty club; of these, only Japan climbed more than ten spots to rise from outside the top-thirty.⁴⁴

Reaching the top echelon of anticorruption ratings need not be policy-makers’ objective. Even shifting from the bottom group of countries to the middle of the pack is a challenging proposition, but one worthy of pursuit. Over the twenty-year period covered by the Control of Corruption indicator, only eight countries in the entire 214-nation data set underwent a 20 percent improvement in their score relative to the range of the indicator (that is, a more than 1-point gain on the −2.5 to 2.5 range; somewhat arbitrarily, I will posit here that this size gain is a sign of an equilibrium shift).⁴⁵Figure 1 shows the top two performers, Georgia and Rwanda. Japan is also shown, because in addition to its move to the top echelon, it is one of the few large countries (with a population of more than one hundred million) to show statistically significant gains over the period.⁴⁶ Despite the three countries’ vastly different histories, demographic sizes, and wealth, together their experiences provide some guidance about why progress across the full accountability equation [a=(t+o+s)*(e-d)] is essential to an equilibrium shift.

Georgia has made the most remarkable anticorruption gains of the last twenty years as it moved from Soviet domination to a post – Cold War period of kleptocracy and finally to a cleaner equilibrium by 2010, when Transparency International ranked the country the most effective in the world at fighting corruption.⁴⁷ These gains were made possible in part by structural effects that may not be relevant to other countries: an educated population, historic resistance to Soviet rule, a head start in removing the economic distortions of Soviet rule via reforms that were already underway by the mid-1990s, and the nearly complete turnover of elites between the Soviet collapse in 1991 and the Rose Revolution in 2003. Contingent political factors also played a role: discontent with corruption, poor public services, and the manipulation of elections contributed to the Rose Revolution, which helped usher in a new generation of policy-makers.

Although these factors may have been helpful, none guaranteed improvement, and comprehensive reforms that fit the accountability equation appear to have played a more central role in altering the corruption equilibrium. Following the 2004 inauguration of President Mikheil Saakashvili, parallel reforms in distinct areas were driven forward with an overarching focus on improving service provision and fighting corruption. Together, this strategy contributed to a “mental revolution” in Georgian society.⁴⁸ Below, I list the confluence of factors that fit within the anticorruption equation outlined above.

• Transparency reforms included efforts to improve the transparency of competitive civil service exams, make public and civil registries work better, automate public service provision to reduce the number of interactions between civil servants and the public, and introduce a competitive common entrance exam to overcome corruption in university admissions.

• Oversight reforms included the decentralization of municipal services, introduction of local elections for mayors, improved monitoring of local governments, and location-specific innovations, such as the use of closed-circuit televisions in university testing centers.⁴⁹

• Sanctions against corrupt officeholders were severe, with many removed from office while the new government focused on exemplary high-level arrests and prosecutions of key officials. Most notably, early in the Saakashvili administration, sixteen thousand traffic officers were fired overnight.⁵⁰

• Institutional effectiveness was improved through competitive hiring practices and salary hikes in the public sector, a focus on improving revenue collection and broadening the tax base, increased collection rates for public utilities, and a reduction in licensing requirements and tariffs. New statutes introduced plea bargaining and asset seizures.⁵¹ New bodies were created to combat corruption, such as the Interagency Council for Combating Corruption, the Internal Affairs Ministry's Anti-Corruption Department, the State Audit Office, and the State Procurement and Competition Agency.

• Although political domination has been high at times – for example, under Saakashvili's United National Movement – the strong consensus in favor of anticorruption reform blunted the potentially damaging effects of political domination in the short term. Eventually, domination led to complaints – which have never been entirely silenced – about hyper-centralization, increasingly authoritarian tendencies, and weak judicial independence.⁵² Partly as a consequence of these concerns, significant political turnover occurred, such as the 2012 victory of the Georgian Dream coalition over the United National Movement.

Despite its gains, Georgia is not Denmark. Tensions in Abhkhazia and South Ossetia and the military conflict with Russia in 2008 permitted areas of intense criminality to emerge.⁵³ Writing in 2009, political scientist Alexandre Kukhianidze noted that the police were still used for political ends, courts remained dependent, human rights were upheld in the breach, and law enforcement was timid in prosecuting elite corruption. But today's Georgia is a stunning success in contrast with the Georgia of two decades prior, when corruption was embedded, organized crime threatened national security, individuals distrusted the government, and the country was even considered a “failed state.”⁵⁴ The nation clearly achieved an equilibrium shift from a pattern of distrust, widespread government abuses, and corruption in both its petty and grand modalities to a new equilibrium in which universalism and impartiality are the default expectation, albeit incompletely realized. This equilibrium shift was engendered by a series of complementary innovations across the entire accountability equation: transparency, oversight, sanction, institutional effectiveness, and political dominance.

Despite its relative poverty, Rwanda has risen in little more than a decade (between 2005 and 2016) from eighty-third to fiftieth place on the Corruption Perceptions Index. Its story – summarized for space's sake here – broadly aligns with that of Georgia. Under bold and hard-nosed leadership, similar to that exercised in Georgia, new transparency reforms were introduced, including whistleblower mechanisms and annual asset declarations for high-level officials.⁵⁵ Oversight was increased through creation of new accountability institutions, including an Ombuds Office, a National Tender Board, and an Auditor General's Office. The National Anticorruption Advisory Council created in 2004 coordinates the efforts of multiple institutions. Regular audits have enhanced oversight and several high-level officials have been prosecuted for corruption-related malfeasance. Government effectiveness is high for the region. Civil society is deeply engaged, in part due to the strong consensus that emerged from the genocide and efforts at achieving postconflict justice and reconciliation.⁵⁶

Two issues remain problematic: first, political domination, including President Paul Kagame's seeming political permanence; second, and partly in consequence, limited judicial independence, which raises questions about the politicization of sanctions. Nonetheless, there has been a significant shift since the 1990s, which has led Rwanda to a new, if still imperfect, equilibrium. The country is now recognized as an anticorruption standout in Africa, and one can imagine that if political domination lessened and the mantle of anticorruption were passed on to new leaders, the new equilibrium might become even more self-sustaining.

Finally, Japan offers an example of just how the process of lessening political domination can help to improve significantly the performance of the full accountability system. After nearly four decades of rule by the Liberal Democratic Party, Prime Minister Morihiro Hosokawa was elected in 1993. His election triggered more effective use of extant accountability capacity while also helping to catalyze both electoral and campaign-finance reforms. Following Hosokawa's compulsory resignation in 1994 and a 1998 information-for-sex scandal in the Ministry of Finance and Bank of Japan, new anticorruption regulations were put into place, including new penal laws on bribery, restrictions on illegal proceeds from mediation, and parliamentary ethics rules. Active competition for political office, in other words, strengthened both the effective enforcement of extant laws and the government's willingness to undertake new reforms in response to scandal.

The diverse experiences described above suggest that countries that are successful at converting policy bursts into equilibrium shifts have simultaneously incorporated reforms across the full (t+o+s)*(e-d) accountability equation. Countries that have not achieved the same equilibrium shift, such as Italy, have often been too reliant on herculean efforts by a single body – prosecutors, for example – without a widespread push to generate or improve performance across the broader accountability equation.

In the remainder of this essay, I sketch some ideas for improving the application of the accountability equation through a more deliberate strategic, iterative, and incremental approach. One reason for a strategic approach is that “big bangs,” especially of the multigenerational sort, are hard to create and sustain, and there are no surefire prescriptions for doing so. Incrementalism is less glamorous but more likely to yield lasting improvements : there is considerable evidence in the academic literature on policy reform that “powering through” by force is seldom as effective as ongoing “problem-solving.”⁵⁷ Incrementalism is effective in helping to develop “second-best solutions” tailored to local needs and contexts.⁵⁸

Future equilibrium shifts might be encouraged or accelerated through a more strategic approach to incrementally identifying and dismantling bottlenecks to accountability. A great deal of strategic thinking is already done at a national level, whether through Transparency International's National Integrity System reviews, the Organisation for Economic Co-operation and Development's Integrity Review process, the National Anti-Corruption Strategies developed under the United Nations Convention Against Corruption, or the Open Government Partnership's National Action Plans.⁵⁹ But even in these national cases, a deeper strategic approach might be helpful, melding: 1) continuous reappraisal of the bottlenecks to transparency, oversight, sanction, and institutional effectiveness; 2) loose coordination of both reappraisal and reform efforts; 3) widespread participation by actors from distinct agencies and civil-society organizations tasked with the component elements of the accountability equation; and 4) the adoption of new tactics in a conscientious and iterative manner.⁶⁰

The marginal effects of anticorruption policy bursts often diminish over time. Progressing from a policy burst to an equilibrium shift therefore requires ongoing and consistently renewed efforts to identify and remedy the most immediately important constraints to accountability. This can prevent backsliding on reforms that have already been undertaken. The path toward more effective accountability thus may often be a matter of identifying key bottlenecks, then brainstorming ways of applying international best practices that both address the bottleneck and fit the local context. How best can countries overcome the particular circumstances that engender the bottleneck? Is it a consequence of policy, legislation, capacity, or process? Particular corruption scandals or court cases may be useful starting points for thinking about which limits on accountability ought to be targeted, drawing public attention to their costs, and building the consensus needed to overcome resistance to reform. More important than any single intervention is regular, structured assessment aimed at identifying the next bottleneck to accountability in a progressive fashion, then prioritizing it and focusing scarce political resources on it in a sustained manner.

Doing so requires that the selection of these bottlenecks be undertaken in a much more analytical and considered fashion than is frequently the case. The strategic discussion of the bottlenecks would ideally proceed stepwise, addressing first the issues that have the largest downstream effects at that particular moment in time.⁶¹ This approach argues for an incremental and iterative attack: it is simply not possible simultaneously to remove all barriers to accountability, identify ex ante the interactions between reforms and existing accountability capacity, forecast the sort of political opposition that may arise, or ensure that all the bottlenecks that might emerge in the future are in fact being correctly identified ahead of time. Incrementalism need not mean glacial reform; it does mean, however, greater attention to problem-solving that weds international best practices with local solutions that are sensitive to political conditions on the ground.

For this reason, it makes sense for reformers to attack the bottlenecks that have the biggest impacts today, evaluate their downstream effect after implementation, and return next year for a new discussion of where the largest impediments to effective accountability are arising in the new, postreform context. As noted earlier, Brazilian society over the past thirty years has proceeded stepwise to enhance transparency, oversight, and sanction.⁶² Most recently, after prosecutorial weaknesses were addressed through organizational innovation (such as the creation of joint task forces) and new laws (such as plea bargaining and antiracketeering laws), a new bottleneck emerged in Brazil: foro privilegiado, the high court's original jurisdiction in cases involving sitting federal politicians. Political defendants are automatically tried in the Supreme Court; due to the Court's dysfunction and overloaded docket, this had become a de facto guarantee of impunity. This bottleneck had long existed, but had not become a salient problem until recently because neither police nor prosecutors were able to effectively bring high-level cases to court at all, let alone push them through the Supreme Court. As more effective investigation and prosecution turned up clear evidence implicating scores of senior politicians in corruption, the foro privilegiado has clearly become one of Brazil's central roadblocks to full accountability. Under public pressure, both Congress and the high court now appear to be considering ways of tackling this newly salient bottleneck.

A central lesson from past experiences of successful equilibrium shifts is that anticorruption gains often occur not merely because of political will or specific policy changes, but as the result of a pattern of continuous reform. Public administration scholar Jin-Wook Choi, for example, demonstrates that the creation of acas – which are often the number-one prescription for corruption – was only the starting point for improvement in Hong Kong and Singapore. Upon their creation, acas were made accountable to the executive, the legislature, and the public. Subsequent reforms then turned to the civil service and to enhancing broader government effectiveness.⁶³ Although acas are often given credit for Singapore's and Hong Kong's gains, neither country stopped at the aca, and subsequent development of the broader accountability system was essential to shifting the corruption equilibrium. Sadly, many anticorruption policy-makers seem only to have recognized the first step – adoption of acas – as the solution. More than 150 countries now have acas, but unsurprisingly, these bodies have produced meager results in all but a handful of cases.⁶⁴

A loose coordinating mechanism may improve the process of strategically selecting and continually tackling bottlenecks. In the United States during the Progressive Era, this mechanism was an active press, working together with civil society. In Brazil, for more than a decade, federal bureaucrats led the effort, working through an annual intrabureaucratic process known as the National Strategy Against Money Laundering and Corruption to come up with reform ideas and carry them forward.⁶⁵ In Hong Kong, the key coordinator has been the ACA. But coordination does not mean control, and it should probably not be too closely identified with a particular politician or political faction for fear that their departure or declining popularity could lead to the collapse of anticorruption coordination efforts.

For this reason, broad participation by relevant accountability bodies in the reap – praisal and reform process is as important as any one specific policy reform. One reason is that on-the-ground expertise will be useful. Another is that the process of uncovering bottlenecks may engender long-term buy-in and create a mental map for reformers across a wide range of bureaucratic agencies to follow in order to engage in the continuous collective action required for reform. A third reason is that simply being asked regularly to focus on accountability keeps it high on the priority list for busy policy-makers.

The accountability equation described here sits on the ladder of abstraction between tactical choices and policy tools (such as procurement reform or antiracketeering laws) and more conceptual governance orders (such as open-access orders or “rule-of-law regimes”).⁶⁶ The equation provides guidance on how to structure tactical policy choices, as well as how it may be possible to move conscientiously and proactively from pernicious cycles of corruption to improved governance.

This framework has several potential benefits. One is that, by focusing on accountability, it expands the deliverables of reform beyond anticorruption alone, providing more incentives for society to undertake the collective action needed to combat corruption effectively.⁶⁷ Reform can make leaders’ jobs easier. After all, anti-corruption need not be the primary selling point for the preservation of a free media, independent courts, active accounting tribunals, or other accountability-enhancing bodies. A free press, for example, does not only report on scandal; it can also provide information to the regime, disseminate official data, and publicize politicians’ achievements; three contributions even the most risk-averse politician may cherish. Independent courts may not only convict corrupt officials, but also uphold (and thereby legitimize) the current incumbent's laws and restrain future coalitions seeking to overturn today's preferred policies. Accounting agencies serve not only to uncover corrupt dealings, but also to evaluate and corral bureaucracies, ensure the effective use of public monies, and see to it that policy goals are being met. In other words, the more deliverables can emerge from accountability agencies, the more likely stakeholders are to preserve and improve them.

Emphasizing the broad accountability equation over the implementation of boilerplate anticorruption “best practices” may also be preferable because it is hard to know where challenges – such as particular forms of corruption, ineffectiveness, or inefficiency in the provision of public policies – might arise. It is therefore difficult to design a single intervention that is certain to “be in the right place at the right time” or to address all of corruption's multifarious forms.⁶⁸ Furthermore, because any single institution can be co-opted or may have unpredictable performance in particular contexts, a broader accountability approach can help ensure that any single agency or reform initiative is buttressed by others.

The accountability equation approach may also be particularly conducive to a long-term strategy of institutional capacity building, whereby islands of excellence are created independently of each other but are slowly patched together by connections that enhance their joint effectiveness. This can offer a partial antidote to problems of agency and leadership. While good leaders always play an important role, anticorruption efforts may be more likely to survive and accumulate when power is dispersed across multiple agencies, leaders, and processes, in such a way that when one falters, others are able to pick up the reins. Finally, an iterative approach across the full accountability framework may enable reformers to sequence reforms in ways that do not directly challenge political elites, or do so only under politically advantageous conditions.