Corruption in any country is an unacceptable tax on economic growth, and an obstacle to the private sector investment needed to reduce poverty. Accordingly, the Millennium Challenge Corporation (MCC) seeks out partner countries that are committed to combatting corruption. Since corruption is complicated and multifaceted, MCC’s approach to understanding and monitoring the issue relies both on the best data and assessments available, and on constantly leveraging a network of experts. This Report to Congress explores in depth the Control of Corruption indicator (and related data) MCC uses, and describes our efforts to strengthen the indicator by 1) consulting the best data available today and 2) working with other stakeholders to improve the quality of data available.
MCC has a comprehensive approach to measuring and analyzing a country’s commitment to Control of Corruption, and seeks to continuously improve this approach. One of the biggest achievements over the past two years in this vein has been MCC’s work with other producers and users of governance data to establish the Governance Data Alliance (GDA). This new initiative brings together civil society, the private sector and the donor community in an effort to bring greater public transparency to the wealth of governance data that is produced, share best practices, and help fill gaps in measuring governance. The GDA recently launched a new, publicly available online dashboard that allows users to see how countries perform on a wide range of governance indices over time, and to compare different countries against each other. Through these and other efforts, MCC is working to bring greater transparency, quality, and efficiency to the world of governance data, and to empower those who share our commitment to fighting corruption.
Why Control of Corruption Matters
MCC works with relatively well-governed, poor countries. In order to assess whether countries are relatively well-governed, MCC creates annual scorecards that use objective and quantifiable indicators in three broad policy categories: Ruling Justly, Investing in People, and Encouraging Economic Freedom. These policy indicators are developed by independent third-party institutions in an analytically rigorous manner. When it comes to corruption, countries must score in the top half of their World Bank-determined income group on the Control of Corruption indicator to pass the scorecard. This makes the Control of Corruption indicator a “hard hurdle,” which places an assessment of anti-corruption efforts at the center of MCC’s eligibility decision-making process.
The inclusion of the Control of Corruption indicator as a hard hurdle is tied directly to MCC’s mission to reduce poverty by promoting economic growth. Economic literature shows the importance of controlling corruption for promoting economic growth and reducing poverty. Corruption hinders economic growth by increasing costs, lowering productivity, discouraging investment, and reducing confidence in public institutions. It can limit the development of small- and medium-sized enterprises, weaken systems of public financial management, and undermine investments in health and education 1 . Corruption can also result in public officials siphoning funds away from essential public services, add a higher level of risk to the investment decisions of low-income individuals, and reinforce patterns of unequal asset ownership – thereby limiting the ability of the poor to borrow, invest, and increase their income 2 .
How MCC measures Control of Corruption
The Control of Corruption indicator on MCC’s scorecard comes from the Worldwide Governance Indicators 3 (WGI). This indicator, produced annually by Brookings and the World Bank, measures the extent to which public power is exercised for private gain (including both petty and grand forms of corruption), as well as capture of the state by elites and private interests. It also measures the strength and effectiveness of a country’s policy and institutional framework at preventing and combatting corruption 4 .
WGI’s Control of Corruption indicator itself is an aggregation of up to 22 different sub-sources, all of which come from third parties. These sub-sources include a wide variety of data, including perceptions surveys, expert and donor assessments, and private sector risk analyses. WGI pulls specific aspects or questions from each sub-source, and then uses a sophisticated weighting methodology to create an aggregate score for each country. Full details are provided in Appendix I.
The WGI Control of Corruption Indicator
For any indicator that MCC uses on its scorecard, we assess its relative strengths and weaknesses by looking through five lenses:
- Availability: Is the indicator publicly available, and produced by a third party?
- Coverage and Frequency: Is the indicator available for the widest possible array of countries, and is it regularly updated?
- Comprehensiveness: Is the indicator looking comprehensively at the issue in question, or is it narrowly focused on only a few specific aspects?
- Methodology: Is the methodology used for creating the indicator transparent and clear, and does it rely to the maximum extent possible on objective versus subjective evidence?
- Actionability: Can a country take clear action, like policy reform or other steps, to improve its performance?
WGI’s Control of Corruption indicator is very strong when viewed through these lenses, though it does have some weaknesses:
- Availability: WGI’s indicator is readily available to anyone at www.govindicators.org, it is produced by an objective third-party (the multilateral World Bank, together with the Brookings Institution think tank), and it uses a broad array of sub-sources that cover a wide spectrum of perspectives and methodological approaches.
- Coverage and Frequency: WGI’s indicator covers all countries in the world (i.e., all MCC candidates have a score), and is updated annually in September. That said, as with any dataset, the indicator does have a data lag, in this case of roughly 1-1.5 years (i.e., the scores produced in September 2015, which MCC used on our November 2015 scorecards for selection for 2016, are largely reflecting events and assessments from 2014 and early 2015).
- Comprehensiveness: The wide range of sub-sources WGI uses ensures that a complex, multifaceted issue like corruption is assessed in all its forms, and in all possible ways.
- Methodology: While the sub-sources used and aggregation methodology are clear (if mathematically complex), the sub-sources themselves are a mix of both transparent, evidence-based narratives and surveys, and some opaque assessments that offer little insight into the specific evidence or process used to create the sub-source. Therefore some sub-sources are more insightful than others.
- Actionability: Because corruption is so multi-faceted and complex, taking concrete policy steps to “improve” a score can be difficult, compared to other scorecard indicators like Fiscal Policy or Education Expenditures, which respond more directly to policy action.
WGI is extremely valuable in that we can “unpack” the sub-sources to see what may be driving a particular score for a given country – i.e., is it a perceptions issue, or an expert assessment issue? Is it more about grand corruption or petty corruption? When available, do the expert narrative-based sub-sources cite specific policy actions or events that caused their assessment to move in a certain way?
However, even when we “unpack” the sub-sources, it is often difficult to move the indicator through immediate policy action. Sometimes the sub-sources do not provide much insight into what reforms are needed, while perceptions data do not identify clear steps a government can take. As a result, improving a low score on the Control of Corruption indicator often requires a sustained, multi-year effort that addresses the issue in many ways. It can be done, however, as countries from Cote d’Ivoire to the Philippines have shown.
Other Corruption Data Sources
MCC continues to look for new measures of corruption, but to date, WGI’s Control of Corruption indicator remains best-in-class across the five lenses we consider. Alternative indicators have a significant shortcoming on at least one of these five factors, making them unsuitable substitutes. Consider three well-known examples:
- Transparency International’s Corruption Perceptions Index (CPI) is easily available, and has wide and frequent country coverage, but is not less directly actionable or comprehensive. CPI is a well-known index that comes out annually, and often garners significant media attention. Based on expert opinion surveys, the CPI measures the perceived levels of public sector corruption worldwide. As a result the CPI does very well in terms of availability and coverage, but it does not do as well as WGI’s Control of Corruption indicator in terms of:
- Comprehensiveness: it looks only at perceptions of corruption, and strictly in the public sector. Private sector corruption, elite capture of the state, experiential assessments, and the nuance between petty and grand corruption are either not included, or not easily extracted from it.
- Methodology: while clear, the CPI relies strictly on expert perceptions. There are few deep narratives, experiential surveys, policy assessments, or other concrete items beyond perceptions of corruption.
- Actionability: Because of its narrow focus on perception, it can be difficult for a country to find actionable policy responses to improve its score.
- Global Integrity’s Country Reports are actionable, comprehensive, and methodologically sound, but lack wide country coverage and are not updated sufficiently frequently. Global Integrity’s reports are perhaps the most useful of the corruption-related assessments available. These comprehensive reports provide not only a score, but also a long narrative on 23 specific policy components. Each report “evaluate[s] both anti-corruption legal frameworks and the practical implementation and enforcement of those frameworks, and takes a close look at whether citizens can effectively access and use anti-corruption safeguards.” 5 As a result, the reports are exceedingly actionable, and their narratives lay out a clear methodology backed up with rigorous analytical evidence. The reports clearly outline where a government is strong or weak on corruption, and specify what it needs to do to address specific issues. However, the reports suffer from a very significant weakness:
- Coverage and Frequency: the reports only cover a few countries each year, and are not annually updated (in fact, the last year for new reports was 2013). As a result, if MCC were to use Global Integrity’s reports, we would no longer have an annually updated indicator covering all candidate countries. We rely on the Global Integrity reports as a source of supplemental information, when and if they are available for a given country.
- The International Budget Partnership’s Open Budget Index (OBI) and Survey provides wide country coverage, is frequently updated, and is actionable, but is not comprehensive, focusing on only one narrow aspect of corruption (budget transparency). The OBI comes out roughly every other year, and covers a broad spectrum of countries. Like Global Integrity, it follows a clear and objective evidence-based methodology, and is actionable – countries are provided with specific policy recommendations on how to improve budget transparency, oversight, and citizen participation. However, by only covering one narrow aspect of corruption – budget transparency – the OBI is not a good alternative to WGI’s Control of Corruption indicator since it does not cover other aspects of corruption. That said, MCC relies on the OBI as a source of supplemental information on budget transparency as necessary.
WGI’s Control of Corruption indicator continues to be the best available for MCC’s needs, though it is not without its weaknesses. MCC continues to review and analyze the other corruption assessments available, and uses them to supplement WGI’s Control of Corruption indicator, when relevant and available.
Analyzing and Responding to Performance on the Control of Corruption Indicator
During the period of compact development, a country is examined for re-selection by MCC’s Board every year. This means scorecard performance is scrutinized closely, including performance on the Control of Corruption indicator. MCC then continues to examine scorecard performance during implementation. As a result, whenever there is a concerning event or trend that relates to performance on the Control of Corruption indicator, MCC looks closely at the underlying causes, and crafts a response accordingly.
MCC assesses a country’s performance on WGI’s Control of Corruption indicator using a median system. To “pass” the indicator, a country needs to have a score better than at least half of its income-level peers (above the 50th percentile). Because we use this percentile-based system, sometimes countries that had scored just above the median (passing) may suddenly have their score fall just below the median (failing) for reasons that have nothing to do with policy changes. Countries are compared to their peers, so changes in the global medians could even cause a country whose score has improved to fail, if the global median were to rise by more than that country’s score improved.
As a result, MCC recognizes the possibility that a compact-eligible country, which passed the Control of Corruption indicator when selected, can generally maintain and improve policy performance, but still fail the Control of Corruption indicator (or some other indicators) in a given year due to one or a combination of the following factors: 6
- Graduation from the low income pool (which have lower medians) to the lower middle income pool (which have higher medians);
- Introduction of new sub-indicators or new methodologies that cause a score to fluctuate by altering the data that score is based upon;
- Improvements in the availability or quality of data that cause the score to fluctuate because of the resulting revisions;
- Rising income-group medians (where the income pool’s median goes up by more than a country’s score); and
- Slight declines in performance (where the score moves a small amount that is not statistically significant, or is within the margin of error).
None of these five factors is likely to be serious enough to warrant a suspension or termination of eligibility, because they are usually not indicative of policy actions on the part of the government that would indicate a pattern of actions inconsistent with MCC’s eligibility criteria. Nonetheless, a sudden failure or decline is still cause for concern, especially on the Control of Corruption indicator. Because of this, when a country unexpectedly fails the Control of Corruption indicator (or shows a concerning score decline), MCC will scrutinize the reasons for the decline in three ways:
- First, MCC will look to see if one of the non-policy-related issues outlined above may be the cause;
- Second, MCC will “unpack” the WGI Control of Corruption indicator by examining what the sub-source indicators say about the reasons for the decline; and
- Third, MCC will consult other sources of supplemental information, as well as experts, to gather a real-time sense of what is going on in the country.
If an eligible country does not pass the indicator in a given year or shows a concerning decline, but (following this scrutiny) has not demonstrated a policy reversal or a pattern of actions inconsistent with the eligibility criteria, MCC may ask the country to develop and implement a policy action plan to address the concern(s), and thereby demonstrate its commitment to meeting the eligibility criteria.
Benin in FY 2014 provides an example. Benin unexpectedly failed the Control of Corruption indicator by moving from just above to just below the median. As there was no clear policy-related explanation for the decline, MCC’s Board did not suspend or terminate Benin, but rather decided to not re-select Benin for FY 2014. MCC then asked the Government of Benin to identify and implement a set of concrete actions to demonstrate its commitment to addressing corruption over the course of 2014. Benin’s actions included funding and empowering a new Anti-Corruption Commission, mandating asset declarations by senior government officials, and creating a process to consult civil society during the development of the national budget. While there is no guarantee that these actions will directly move the Control of Corruption score, they are examples of the type of efforts that are indicative of a government taking the serious policy steps needed to address the situation. In FY 2015, Benin showed one of the strongest score improvements among low income countries, and once again passed the Control of Corruption indicator. While it is impossible to show direct causality, it is this type of commitment to the issue which helps lend credibility to the possibility of sustained longer-term improvement on the score. In Benin’s case, it was re-selected in FY 2015, and signed its second compact in September of that year.
Conversely, if it is found that a country which has failed the Control of Corruption indicator has demonstrated a significant policy reversal, MCC may issue a warning, suspension, or termination of eligibility and/or assistance, depending on the specifics of the situation. Even if the country has not yet failed the indicator, but has shown a significant decline in performance concurrent with clear negative policy actions, MCC may still take action.
It is important to note that MCC has formally suspended / terminated assistance or eligibility on 10 occasions, though in all cases to date this choice was made because of severe democratic rights declines, such as coups or stolen elections. There has not yet been a case of an MCC partner showing such a profound decline on the Control of Corruption indicator – or engaging in policy actions that would indicate they were backing away from a commitment to combat corruption – as to warrant a suspension or termination. There have been several examples of countries coming very close, however, as shown with the Benin example above.
Continuous Improvement on Measuring Corruption: The Governance Data Alliance
Beginning in late 2013, a number of major producers and users of governance data – including MCC – came together to discuss the current challenges in governance data, including data on control of corruption. Together, the group identified many challenges in the governance data marketplace, including:
- Coverage: some countries and topics are well covered, others are not;
- Coordination: some efforts are duplicative, and opportunities for efficiencies are lost;
- Need: sometimes the data produced is not necessarily what is needed by the users; conversely, users are sometimes not articulating what it is they need from data producers; and
- Reliability: variable funding streams, differing data collection methods, and other related logistical challenges can make the resulting data unreliable, incomparable or unpredictable over time.
As a result, MCC, together with these other producers and users of governance data, established the Governance Data Alliance (GDA) to begin addressing these challenges. A list of the initial members is provided in Appendix IIa. Ultimately, the GDA set itself four tasks: 7
- Publishing and promoting a governance data dashboard: This effort aggregates the high-quality data from the GDA’s producers in one place. It promotes improved accessibility, standardization, and use of existing governance data.
- Knowledge sharing and collaboration among data producers: This task supports peer exchange and technical assistance between groups within the GDA, as well as interested parties who have not yet joined, on topics including research methodology design, data analysis tools, data storage, data visualization techniques, and managing government feedback.
- Collecting and analyzing governance data user habits: This work stream seeks to better understand and meet the needs of governance data users in future data production efforts. Activities involve an analysis of responses collected through a major 2014 Reform Efforts Survey of governance data users worldwide, as well as the collection of new data through novel “snap polls” focused on understanding future demand for governance data.
- Piloting coordinated data co-production efforts: The goal here is to explore possible economies of scale in data production. The work stream aims to improve the availability of data, particularly for under-assessed countries, and to drive down the costs of data production over time.
In late February 2016, the GDA’s governance data dashboard was launched. It is available at http://www.governancedata.org. The dashboard currently provides over 20 datasets, covering critical governance topics including budget transparency, rule of law, and various aspects of democratic rights. A full list is provided in Appendix IIb. With one click, a user can see how many of these datasets are available for a particular country, and compare the performance of a country with several others. Historical time-series can also be produced to see changes in a country’s performance over time, where and when the data is available. As a result, the dashboard provides users of governance data with a “one stop shop” to see how a country performs and compares on a wide range of governance indices.
MCC has been strongly supportive of the GDA and its efforts to bring greater transparency, quality, and efficiency to the world of governance data. The dashboard will help MCC more effectively and efficiently identify the full array of assessments available for a specific country, particularly on control of corruption. While the dashboard will not (at this point) replace the WGI Control of Corruption indicator, it will greatly augment it. Just as MCC will consult sub-sources, supplemental information, and experts to deeply understand a country’s performance on control of corruption, the GDA adds another tool to that arsenal.
Ultimately, the dashboard and the other GDA work streams will help MCC better understand the nature of corruption and related governance issues in its partner countries. The GDA’s work will also help to ensure that the producers of such data are coordinating in better ways and creating data that is most useful to users like MCC. The dashboard is just one aspect of MCC’s support for efforts to ensure continuous improvement in how corruption is measured and understood.
MCC remains committed to ensuring it works only with partner countries who take combatting corruption seriously. Corruption is a complex topic, with many manifestations. Understanding the nature of corruption in a country, and determining how to address it, requires constant examination and a wide set of analytical tools. From the baseline established by the WGI Control of Corruption indicator itself, to the wider set of other indicators, experts, supplemental information, and the new Governance Data Alliance, MCC has a powerful network to draw on as it seeks to understand and respond to this critical issue in its partner countries. MCC will continue to review and improve this approach as the universe of corruption data innovates and evolves in the coming years.
Appendix I: The sub-sources and specific questions / aspects included in the Worldwide Governance Indicators’ Control of Corruption indicator
Economist Intelligence Unit:
- Assessment of Corruption among public officials
World Economic Forum Global Competitiveness Report:
- Public Trust in Politicians
- Diversion of Public Funds
- Irregular Payments in Export and Import
- Irregular Payments in Public Utilities
- Irregular payments in tax collection
- Irregular Payments in Public Contracts
- Irregular Payments in Judicial Decisions
- State Capture
Gallup World Poll:
- Is corruption in government widespread?
Institutional Profiles Database:
- Level of "petty" corruption between administration and citizen
- Level of corruption between administrations and local businesses
- Level of corruption between administrations and foreign companies
Political Risk Services International Country Risk Guide
- Section on Corruption
Global Insight Business Conditions and Risk Indicators
- Assessment of the risk that individuals/companies will face bribery or other corrupt practices to carry out business, from securing major contracts to being allowed to import/export a small product or obtain everyday paperwork. This threatens a company's ability to operate in a country, or opens it up to legal or regulatory penalties and reputational damage.
African Development Bank Country Policy and Institutional Assessments
- Assessment of transparency, accountability and corruption in the public sector
- How many elected leaders (parliamentarians) do you think are involved in corruption?
- How many judges and magistrates do you think are involved in corruption?
- How many government officials do you think are involved in corruption?
- How many border/tax officials do you think are involved in corruption?
Asian Development Bank Country Policy and Institutional Assessments
- Assessment of transparency, accountability and corruption in the public sector
Business Enterprise Environment Survey
- How common is it for firms to have to pay irregular additional payments to get things done?
- Percentage of total annual sales do firms pay in unofficial payments to public officials?
- How often do firms make extra payments in connection with taxes, customs, and judiciary?
- How problematic is corruption for the growth of your business?
Bertelsmann Transformation Index
- Assessment of anti-corruption policy
- Assessment of prosecution of office abuse
Freedom House’s Countries at the Crossroads report
- Assessment of Anti-Corruption and Transparency
Freedom House’s Freedom in the World report
- Assessment of Corruption
Transparency International’s Global Corruption Barometer Survey
- Frequency of household bribery - paid a bribe to one of the 8/9 services
- Frequency of corruption among public institutions: Political parties
- Frequency of corruption among public institutions: Parliament/Legislature
- Frequency of corruption among public institutions: Media
- Frequency of corruption among public institutions: Legal system/Judiciary
- Frequency of corruption among public institutions: Public officials
Global Integrity Index
- Assessment of accountability
IFAD Rural Sector Performance Assessments
- Assessment of accountability, transparency and corruption in rural areas
- Assessment of frequency of corruption
World Bank Country Policy and Institutional Assessments
- Assessment of transparency, accountability and corruption in the public sector
Political Economic Risk Consultancy Corruption in Asia Survey
- Question on to what extent does corruption exist in a way that detracts from the business environment for foreign companies?
Vanderbilt University Americas Barometer
- Frequency of corruption among government officials
Institute for Management and Development World Competitiveness Yearbook
- Bribing and corruption exist in the economy
World Justice Project
- Assessment on the Absence of Corruption
Appendix IIa: Current members of the Governance Data Alliance 8
- Amida Tech
- Centre for Law and Democracy
- Freedom House
- Global Financial Integrity
- Global Integrity
- Hewlett Foundation
- International Budget Partnership
- Millennium Challenge Corporation
- Natural Resource Governance Institute
- Omidyar Network
- Open Government Partnership
- Organization for Economic Cooperation and Development
- Programa Nacional de la Competitividad
- Results for Development
- Transparency International
- World Bank
- World Justice Project
- World Resources Institute
Appendix IIb: Current datasets available on the the Governance Data Alliance dashboard 9
- African Integrity Indicators (from: Global Integrity)
- Citizen Engagement in Rule-Making (from: World Bank)
- Corruption Perceptions Index (from: Transparency International)
- Doing Business – Distance to the Frontier (from: World Bank)
- Doing Business – Ranking (from: World Bank)
- Environment Democracy Index (from: World Resources Institute)
- Freedom in the World (from: Freedom House)
- Freedom of the Press (from: Freedom House)
- Freedom on the Net (from: Freedom House)
- Global Integrity Report (from: Global Integrity)
- Hot Money Narrow Outflows (from: Global Financial Integrity)
- Illicit Financial Flows (from: Global Financial Integrity)
- Money, Politics, and Transparency (from: Global Integrity)
- Open Government Partnership Commitments (from: Open Government Partnership)
- Open Budget Index (from: International Budget Partnership)
- Open Government Index (from: World Justice Project)
- Right to Information Rating (from: Centre for Law and Democracy)
- Regular Forum for Open Government Consultation (from: Open Government Partnership)
- Resource Governance Index (Natural Resource Governance Institute)
- Rule of Law Index (from: World Justice Project)
- Trade Misinvoicing Outflows (from: Global Financial Integrity)