Engaging corruption: new ideas for the International Monetary Fund

Abstract The IMF’s standard ways of engaging with countries need redoing for complicated, politically loaded issues like corruption. This paper suggests what might be done inside the Fund and with client countries to enable more effective policies. The proposed methods could lead to a broader rethink of how the Fund does business – and they have more general applications to international financial institutions and foreign aid.


Introduction
In April 2018, the International Monetary Fund's Executive Board endorsed "a new framework for stepping up engagement on governance and corruption in our member countries" (International Monetary Fund 2018a). The IMF's Managing Director Christine Lagarde explained, "We have now adopted a framework for enhanced engagement on governance and corruption that aims for a more systematic, evenhanded, effective, and candid engagement with member countries" (Lagarde 2018).
She put four important words in italics. This article emphasizes another: engagement. On complex, sensitive issues like corruption, engagement must go beyond an external appraisal and the importation of external norms and practices. Ideally, engagement means sharing local and international knowledge and jointly developing creative, practical strategiesand then implementing them together.
Engagement is especially difficult regarding corruption and governance. Sensitivities and vested interests lurk. Leaders and officials may experience conflicting cultural normsfor example, impartial and efficient public service versus taking care of and favoring family, clan, and friendship networks. A kind of equilibrium of corruption can emerge: even people whose personal ethics are exemplary are expected, perhaps even required, to participate.
1. Assemble data on governance and corruption in the six sectors identified in the Fund's new framework. 1 Use such data to show the various kinds of economic and social costs that are linked to poor governance and increased levels of corruption (Hlatshwayo et al. 2018;Jarmuzek and Lybek 2018; IMF 2019). 2. Identify in the Fund's experience success stories of improvements in each sector and then develop implementation-oriented case studies of them. 2 3. Pull together economics-based frameworks for reform in each of the six sectors, with the goal of redirecting sensitive perceptions from culture and ethics to structures and incentives.
This article shows how this process might be applied in Ukraine. It proposes that the Fund begin internally, using convenings of IMF staff to consider the "how-to" of helping countries fight corruption even more effectively.

Two notions of the expert
Contrast two approaches. One is "the expert" vis-a-vis the patient. The scientist, the doctor, the therapist, the IMF economists know what works. How? From validated theory and empirical evidence about the patient's situation. The expert's job is to appraise the patient, apply the theory, and prescribe the treatment. The patient's job is to comply and adhere to the treatment.
Needless to say, sovereign nations (not to mention patients) do not enjoy being told what to do. In international development, the expert model above is severely strained (Easterly 2014).
What to do? A second notion says that both outsiders and insiders have expertise, and the task is to combine forces and catalyze creative analysis and action. Locals have tacit and explicit knowledge about local settings. But they may lack access to some of the best data and techniques of data collection to calibrate their challenges. They may not be familiar with examples of what has worked elsewhere, not so much to copy as to inspire. They may benefit from exposure to models and theories that can be used to suggest frameworks for policy analysis. The challenge becomes how to combine forces: how to bring what they know best (local contexts and possibilities) with what outside experts may offer (facts, examples, frameworks).

What is "convening"?
Convening tries to bring together these capabilities in a safe space through a pragmatic process. Those convened have different if overlapping objectives, different if sometimes overlapping capabilities, and different if overlapping information about the state of the world and about if-then relationships. The stakeholders are strategically connected, in the sense that what one party does often affects the outcomes of what other parties do. They are not fully aware of each other's objectives, capabilities, or information sets; they do not fully understand their strategic interrelations. Convening brings together their strengths and inspires them to address their challenges with new information, examples, and frameworks.
The kinds of convenings recommended here provide stakeholders with: 1. Data, especially data that help people "get on the same page" about the nature of the problems. Data-rich discussions help build trust, particularly about controversial issues. The data help everyone to contextualize the challenges. 2. An example of success in similar problem areas, which spotlight what was done and how. The example is conveyed through a part A, part B case study. Participants analyze the issues and make recommendations, then learn what happened, then discuss. The success story from another country provides inspiration. 3. A framework for understanding goals, alternatives, if-then relationships, and strategic interdependence. The framework can provide a new way of conceptualizing the problem. The framework conveys principles. 4. An imaginary news story of success five years hence. Participants read it aloud, then ponder what steps could lead from now to then. The imaginary news story stimulates creative problem-solving.
Convening is one of a number of tools for policy design, including "problem-driven iterative adaptation" Woolcock 2013, 2017), "the learning spiral" (Blindenbacher 2010), "development evaluation" (Patton 2011), and "multistakeholder dialogues" (Susskind, Fuller, Ferenz and Fairman 2003). The website Participedia.net offers a cornucopia of different techniques for different situations. 3 What distinguishes convening as I describe it here is that it builds on policy analysis and evaluation by combining local knowledge and international expertise.

An example of convening about corruption 4
In June 2010, Benigno Aquino III was elected President of the Philippines. His campaign slogan was Kung walang korap, walang mahirap -"When no one is corrupt, no one will be poor." His inaugural address focused on fighting corruption: During the campaign, we said, "If no one is corrupt, no one will be poor." That is no mere slogan for postersit is the defining principle that will serve as the foundation of our administration.
Our foremost duty is to lift the nation from poverty through honest and effective governance … No more influence-peddling, no more patronage politics, no more stealing. No more sirens, no more short cuts, no more bribes (Official Gazette 2010).
When Aquino took office, the Philippines did not fare well on various measures of corruption and competitiveness.
Transparency International's Corruption Perceptions Index 2010 gave the Philippines a score of 2.4 on a 10-point scale, ranking it 133rd out of 178 countries, tied with Azerbaijan, Bangladesh, Honduras, Nigeria, Sierra Leone, Togo, Ukraine, and Zimbabwe. In the 2009-10 Global Competitiveness Index compiled by the World Economic Forum, the Philippines, with a score of 3.9, ranked 87th out of 132 countries, trailing behind countries such as Gambia and Guatemala. In 2010, the International Finance Corporation ranked the Philippines 144th of 183 countries in terms of "ease of doing business", behind the Syrian Arab Republic, Mozambique, and Paraguay.
Several weeks after Aquino took office, in August 2010 an all-day Cabinet meeting was convened in Malacañang Palace. The participants were the new cabinet secretaries, as well as the heads of the Central Bank, the Customs Bureau, and the Bureau of Internal Revenue. Their goal was to develop a practical strategy across the ministries and, indeed, across the public-private divide.
During that Saturday, the participants studied data about the extent and social costs of various kinds of corruption in the Philippines and elsewhere. In the style of a two-part business school case, they analyzed a country that had successfully reduced corruption. They worked through economic models of corruption and a framework for policy analysis.
And then, thus stimulated, they analyzed the situation of their own country. The problems of corruption are complex; the Philippines' setting is unique. And yet, the data, the case study, and the models became, with these participants in this process, a catalyst for creative problem-solving. By eight o'clock that evening, they had devised an outline of a national strategy, which three ministers developed on Sunday and Monday and briefed to the President on Tuesday. A plan of action followed.
As President Aquino's intrepid anti-corruption campaign developed, it included identifying and punishing some major offenders, forging new partnerships with business and civil society to analyze corrupt systems and improve accountability, using citizen scorecards to gauge the performance of government agencies, implementing radical reforms in bottom-up budgeting and evaluation, enhancing coordination across key government agencies, and more. International donors and lenders played key roles in supporting many of these initiatives.
The results have been remarkable: In The Philippines still has a long way to go, but progress in the Philippines under President Aquino was tangible. Even in a country plagued by systemic corruption, improvement was possible, with significant developmental benefits.

A suggestive example: Ukraine and the IMF
How might the IMF and a partner country create a convening about corruption and governance? Simply as an illustration, let the country be Ukraine and the focus be on the financial and banking sectors.

Data
Assembling and sharing data is helps all parties "locate" the challenges in an international perspective. For this article, I have assembled a variety of data on the health of a country's financial and banking sectors, as well as some economically relevant measures of corruption and governance. Appendix 1 describes the variables; they do not include privileged data that may be possessed by the IMF. Appendix 1 also summarizes a data-reduction exercise that extracts two composite measures of the quality of governance in a country's financial and banking sector.
Appendix 2 outlines how this information might best be shared in a country, using the example of Ukraine. It is advisable to begin with the positive. "Along these variables, Ukraine is quite competitive internationally." Then ask the audience, "Given these positives, what could be holding back investment and financing?" And then show the dimensions along which Ukraine is not at all competitive.
At this point, the data on the financial sector can be examined. How does Ukraine fare on various measures?
When the data have been presented, a discussion follows. What are the implications? Sometimes issues of validity and reliability are raised. But experience suggests a major outcome of considering such data together: "We have problems here that are holding our country back."

Success stories
When policymakers work together through an example of progress in another country, their responses are often surprise, insight, and new motivation. The choice and presentation of the case are crucial. One needs a deep understanding of the challenges faced by leaders of Ukraine and intimate knowledge of how another country made progress in dealing with similar challenges, ideally with the IMF's assistance.
How the case is presented is important. I recommend a "live case," which requires no advance readinga good thing with busy leaders. The case is divided into two parts. Part A presents, in perhaps 15 slides, the challenges faced by policymakers in that other country.
Then participants are then divided into groups of four and asked to spend 45 minutes together addressing three questions: (a) what are the major policy problems in the case? (b) what are the alternatives for dealing with them? and (c) what do you recommend? One member of each group is designated as the facilitator and another as the rapporteur.
After a break, the rapporteurs share each group's answers to (c). Then Part B presents what the country's leaders actually did and how, including lurches and midcourse corrections. It also describes the results, including improvements in governance and the effects on investment, political perceptions, and other salient outcomes.
A general discussion follows. Sometimes, the first point made is "That country is not the same as our country." But quickly follow observations such as "But it's interesting how they did this first and not that" or "Look how they neutralized this corrupt force" or "Notice how they built on quick wins to create political momentum." Ideally, the case shows how the IMF's partnering with the country made a difference. For example, the Fund's support may have enabled expenditures outside the constraints of the country's budget. Technical assistance may have been crucial. And the Fund may have helped the country credibly commit to policy reforms, easing implementation and attracting investors.
Clearly, the choice of success stories, sophistication in creating live cases, and skill in managing a creative discussion are crucial. More on these points below.

Framework
Corruption is a sensitive issue because it is so readily linked to ethics, character, and political culture. It is therefore both a relief and a revelation to policymakers to recast corruption as a problem of structures, incentives, and information.
For example, a simple formula such as Corruption ¼ Monopoly þ Discretion -Accountability may help people rethink ways to reduce corruption by reducing monopoly power, clarifying or limiting discretion, 5 and enhancing accountability in a variety of ways (Klitgaard 1988).
Corruption is a crime of calculation, not of passion. A simple decision tree shows how the calculation by the giver and the receiver of bribes is a function of the probability of being caught and convicted, the penalty, the salary, the size of the bribe, and the self-assessed "moral costs" associated with the bribe. The size of the bribe depends on market conditions. In some cases, when bribery becomes widespread an equilibrium results that is analogous to a Prisoners' Dilemma. Corrupt systems themselves can be analyzed: how do they work, where are they weak, and how might they evolve or retaliate? For each kind of corruption, who is hurt, who is helped, and how big is the social cost (including distorted policies and the erosion of trust)? What might be done about each kind? With what benefits and costs (in money and political currencies) for each anti-corruption initiative? Indeed, a kind of checklist might be derived and shared, not as the end of a discussion but the beginning to one. "How does such a model help us identify weaknesses in our systems, to analyze corrupt systems, and think about practical ways to overcome them?" (Klitgaard 2015).
Ideally, this framework would echo themes raised in the success story the group has addressed. "Note how that country's reforms were consistent with these points."

An imaginary news story
The next step is optional but can be remarkably effective: an imaginary news story five years hence of country X's own success in improving governance in the banking sector.
Participants read the news story aloud, a different person taking each line. At the end, there may be applause. Then participants are asked, "What chain of events might lead to such a happy story?" Each participant spends about ten minutes writing down some ideas. Then participants are paired. Each is asked to share his or her ideas with the partner, with the knowledge on both sides that the partner will report to the group what the partner thought was his or her best idea. Then the roles are reversed.
At the end, each person reports his or her partner's best idea, and the facilitator writes all the ideas on a white board.
Then the group is asked, "What together could you do in the next six months to make these good ideas more likely to occur and succeed?" The results are often extraordinary. As noted, one outcome in the Philippines was the formation of that three-person ministerial task force to take these ideas and develop an action plan for eventual discussion and approval.

Convening inside the International Monetary Fund
How might these same ideas might be explored within the IMF? Imagine a convening of IMF staff working on a particular area of governance and corruption.
1. Data. Staff members would together consider the best data to use to characterize a country's competitiveness and the governance of its financial sector. (Similar questions apply in the other sectors in which the Fund might be active in fighting corruption and improving governance.) 2. Success stories. Staff would surface and discuss examples of successful reforms in this area, especially those in which the IMF was involved. "Success" does not mean the eradication of corruption; rather, as in the Philippines under Aquino, it refers to progress with economically relevant outcomes. What were the IMF's roles in enabling these successes? How might success stories be shared in a practical and engaging fashion inside the Fund itself? 3. Frameworks. Staff would discuss economic frameworks and policy-relevant checklists regarding the selected area of governance. 4. Imagining success. Staff would consider an illustrative hypothetical news story, and learn how this device can be used to motivate creativity. 5. Discussion of uses within the Fund and with clients.

Extensions
The International Monetary Fund has long been criticized for its interactions with client countries (Masters and Chatzky 2018). The issues, of course, go beyond the IMF to many forms of foreign aid, where some recipients decry a cookie-cutter approach. Sympathetic observers recognize that when the stakes are high, formal criteria and highly prepared loan packages are essential. Under conditions of macroeconomic crisis, the Fund's authority may depend on importing "solutions" based on tested models that have been applied to many countries. Indeed, a recipient country may simultaneously condemn the Fund's formulaic prescriptions and use them to solve political economy challenges domestically. 6 But with governance issues, there is no parallel to a macroeconomic model. Moreover, the IMF and the recipient find governance issues even more sensitive than macroeconomic reforms. The temptation is to call for, and settle for, form rather than reform: some new laws, creating an anti-corruption agency, some oversight measures, rather than their harder-to-measure implementation.
Convening offers an alternative. This collaborative process shares data, examples of progress, useful analytical frameworks, and an imaginary news story of future success. The results are new ideas and partnerships that improve traction and outcomes-ideas that neither the IMF nor the country might have come up with on its own.
Corruption is a new priority for the IMF. Convening would be a new method for both internal and external application. If successful, experiments with convening might help the Fund rethink how it engages with partner countries on many other issues, even macroeconomics.

Disclosure statement
No potential conflict of interest was reported by the author. Notes 1. Fiscal governance, financial sector oversight, central bank governance and operations, market regulations, rule of law, and anti-money laundering and combating the financing of terrorism. 2. For example, fill out the five case studies summarized in Akitoby 2018. 3. https://participedia.net/en/browse/methods?f%255B0%255D¼field_completeness%3A5 &f%255B1%255D¼field_completeness%3A4 4. This account is drawn from the author's facilitation of this convening and another one 16 months later. 5. "Discretion" may have different interpretations in different legal traditions. In a more principles-based tradition, everything is typically permitted unless explicitly prohibited. In a more prescriptive tradition, such as the former Soviet Union, everything may be more or less prohibited unless explicitly permitted. 6. The Peruvian President Fernando Bela unde Terry once privately thanked the Organization of American States for insisting on land reform as a condition for aid, before loudly condemning the OAS in front of the legislature. "Don't blame me, blame the imperialists-but we need this aid." His argument worked. 7. Bruno Snow provided able research assistance.
Another indicator was drawn from World Bank (2017). Getting Credit "measures the legal rights of borrowers and lenders with respect to secured transactions through one set of indicators and the reporting of credit information through another. The first set of indicators measures whether certain features that facilitate lending exist within the applicable collateral and bankruptcy laws. The second set measures the coverage, scope, and accessibility of credit information available through credit reporting service providers such as credit bureaus or credit registries." A measure called Z-Score came from the Federal Reserve Bank of St. Louis (2018). This measure captures the probability of default of a country's banking system. Z-score compares the buffer of a country's banking system (capitalization and returns) with the volatility of those returns. The z-score is defined as z (kþm)/r, where k is equity capital as percent of assets, m is returned as percent of assets, and r is standard deviation of return on assets as a proxy for return volatility. A higher z-score therefore implies a lower probability of insolvency. We used the natural log transformation.
Bank cap-assets: Bank Capital to Total Assets (most recent year, 2015-2017) -World Bank Open Data (2018)the ratio of bank capital and reserves to total assets. Capital and reserves include funds contributed by owners, retained earnings, general and special reserves, provisions, and valuation adjustments. Capital includes tier 1 capital (paid-up shares and common stock) and total regulatory capital, which includes several specific types of subordinated debt instruments that need not be repaid if the funds are required to maintain minimum capital levels (these comprise tier 2 and tier 3 capital). Total assets include all nonfinancial and financial assets.
Saved at a financial institution (% age 15þ) (2017) -World Bank Open Data (2018)the percentage of respondents who report saving or setting aside any money at a bank or another type of financial institution in the past 12 months. We used the natural log transformation.
Borrowed from a financial institution (% age 15þ) (2017) -World Bank Open Data (2018) the percentage of respondents who report borrowing any money from a bank or another type of financial institution in the past 12 months. We used the natural log transformation.
These data were factor analyzed (Table A2). Two factors had an eigenvalue of 1 and were therefore retained according to the Kaiser rule. The first factor built strongly on the four variables from the World Economic Forum plus lesser contributions from the three World Bank Variables. It accounts for about 52 percent of the overall variance among these 9 variables.
The second factor picks up the Doing Business measure of Getting Credit.
Interestingly, the first factor is correlated 0.83 with the country's Global Competitiveness Score, itself a composite of 112 variables. In contrast, the second factor is correlated only 0.13 with the Global Competitiveness Score.
LR test: independent vs. saturated: chi2 (36)¼669.29 prob > chi2 ¼ 0.00 Rotated factor loadings (pattern matrix) and unique variances Factor rotation matrix Of the 100 countries in the sample, the ten countries with the lowest scores on Factor 1 were: Lesotho, Greece, Sierra Leone, Malawi, Moldova, Mozambique, Ukraine, Argentina, Madagascar, and Zambia.
The ten countries with the highest scores on Factor 1 were: Singapore, Luxembourg, Finland, USA, Switzerland, Canada, Sweden, Germany, Japan, and Hong Kong China.
Appendix 2. Example of presenting a country's data on governance and corruption Imagine a presentation focused on Ukraine. The presentation begins with the positive, makes comparisons with other countries, and then moves to areas where the country is less competitive in terms of perceptions about governance and corruption.
The presentation uses these data:

Global Competitiveness Report 2017-2018
Overall ranking 81 of 137 countries (41st percentile). Some areas where Ukraine is quite competitive: