Perception aggregated indicators like Corruption Perception Index have low specificity and actionability and even lower sensitivity to change due to their nature of aggregates of experts cores. The 2017 Index for Public Integrity, although more specific, is also not optimal to trace change across time due to its aggregation. After tracing each component of the index separately this paper proposes a trends analysis and a forecast based on disaggregated IPI component rather than the index itself. Using 12 years as a time interval and a three steps method a set of improvers, stationary and declining cases are proposed. The paper also suggests how the new tools should be used by promoters of good governance.
Lagging governance or lagging indicators?
While research of the last twenty years from academia, OECD and Bretton Woods institutions has created more tools to study good governance than ever before, knowledge remains scarce on what triggers institutional evolution. The most popular governance indicators are perception ones, which have proved vulnerable to the reputation loss of countries. In Europe, high deficit countries, most notably Greece, deteriorated on their Corruption Perception Index after the euro crisis hit, exactly when they had embarked on reforms to correct imbalances and rationalize their governance. Once the popular perception attributed their economic problems to their chronic governance quality, experts rushed to correct their previously high rankings, thus punishing precisely the more reform-minded governments, which were trying to fix the problem.
Measurement of institutional change is a formidable problem, and despite advances in corruption measurement due to open data there is little guidance for policymakers. A United Nations Development Programme review report noted on the measurement of corruption: ‘To put it plainly, there is little value in a measurement if it does not tell us what needs to be fixed’ (UNDP 2008, p. 8). We need to establish both the incidence of corruption and its evolution over time if we are to succeed in diagnosing and solving problems of governance or institutional quality in public policy.
While the obstacles have proved important for obvious reasons, such as the hidden nature of corruption (although that varies greatly according to the prevalence of the phenomenon), such obstacles are also over-estimated because of limited communication across disciplines or simply the reluctance of policy actors to open access to publicly derived data. Following the lead of Johnston (2010) who argued that the kind of evidence that we should prioritize should help reformers identify anti-corruption priorities, apply effective countermeasures and track the effects of their efforts we briefly review in this paper the potential of the most powerful corruption determinants to predict change in control of corruption.
Due both to their general definition of corruption as quality of governance and their constitutive process as aggregates, the widely used perception indicators from Transparency International or World Bank have proved rather insensitive to change. Expert scores, as Daniel Treisman (2007) remarks, raise uncertainty as to what sort of lags one should expect before political or economic changes influence perceived corruption (p. 220). Comparing across averages may also be seen as problematic in itself, in particular since the number of sources varies across countries because certain countries have more experts rating them, others fewer, which may feed bias (as in the case of Qatar, consistently rated well in the years when corruption allegations over the World Cup and defence contract bribery were emerging). Furthermore, new sources keep being added and others discontinued (Kaufmann and all 1999). The World Bank researchers Kaufmann and Kraay (2002) calculated transparently that about half the variance over time in the World Bank indexes for Rule of Law and CoC resulted not from evolution of scores within individual sources but from changes in the sources used and the weights assigned to different sources (pp. 13–14). Nevertheless, they later argued that certain changes over longer periods are indeed significant (Kaufmann and all 2007). CoC has a confidence error built in that essentially measures consistency across sources and checks for significant change. However, only a handful of countries evolved to the upper third of the scale, where corruption becomes an exception, and in fact, hardly any did if we remove the Caribbean tax havens (Mungiu-Pippidi and Johnston 2017).
There is not much actionability in such results, even if they might reflect the reality of governance lag. First, too little change exists that we can trust to be real change. Second, the indicators cannot suggest reasons for change or lack of it and are too non-specific to tell us where government fails or where it improves. It is not even clear if what is measured is how corrupt a government is or how corrupt its society is: in fact, it is probably both. Nevertheless, the available results are widely used by international aid donors, chancelleries of various governments and the media. More often than not they are misused, mostly by citing ranks across time for different numbers of countries assessed yearly (with CPI), or interpreting small differences which, given the arbitrariness of scaling, have no significance at all (Andersson and Heywood 2009). Largely, while corruption aggregates play a largely positive role in research and raising awareness of the problem of corruption, they proved less useful as tools with which to guide action.
IPI as innovation and its limitations
Yet the capacity to control corruption and enforce public integrity at the country level could and should be measured, since the national context is both the arena for socializing citizens into the dominant norm of honesty and the site of major policy action (Mungiu-Pippidi 2015; Gächter and Schulz 2016).
Mungiu-Pippidi and Dadašov (2016) offered an attempt to organize the empirical findings of the vast literature predicting corruption at national level to derive a measurement, the Index for Public Integrity (IPI)[1]. Corruption determinants can be divided into either factors that generate corruption –resources or opportunities – or constraints – factors that can hinder corruption (Mungiu-Pippidi 2015, ch. 4). Resources and constraints are to a great extent development dependent (proxied by the Human Development Index, which explains about half the variance in corruption), and after testing their associations and interactions one most effective proxy per category was selected to illustrate the categories, with a principal component then extracted. The selected proxies are listed below, and their association with perception of corruption (with development control) is shown in Table 1.
For resources/opportunities:
Administrative burden measures the extent of domestic bureaucratic regulation, which results in higher risk of corruption. The component is constructed by combining the average number of procedures and the time needed to start a business and pay corporate tax. The data stems from the World Bank’s Doing Business dataset.
Trade openness measures the extent of regulation concerning a country’s external economic activity. The component uses data from the World Bank’s Doing Business datasets to combine the average number of procedures and time taken to export and import goods.
Budget transparency measures to what extent the public has access to documents allowing it to control discretionary public spending. The component is based on selected questions, which are used for the Open Budget Survey provided by the International Budget Partnership.
For constraints:
Judicial independence captures to what extent the judicial system is autonomous from both government and private interests. The data stems from the Global Competitiveness survey developed by the World Economic Forum.
E-citizenship captures the ability of citizens to use online tools and social media to exercise social accountability. The component is constructed by combining the number of broadband subscriptions and internet users with the proportion of Facebook users among the population. The data stems from the International Telecommunication Union and Internet World Stats.
Freedom of the press measures the degree of media independence resulting from a specific national legal, political and economic environment. Free media are indispensable for monitoring democratic institutions, public accountability and good government. The component is a fraction of the Freedom House’s Freedom of the Press report.
Table 1: The predictive power of proxies of the Index for Public Integrity
Legend: OLS regressions. The dependent variable is the WGI CoC 2014. HDI = Human Development Index from the United Nations Development Programme; t statistics in parentheses with * p < 0.05; ** p < 0.01; *** p < 0.001. Robust standard errors are used. N=105.Source: Mungiu-Pippidi and Dadašov (2016)
IPI predicts 87 percent of CPI, 52 percent of the Global Corruption Barometer main question (‘How many officials are corrupt?’ for a universe of 91 countries) and correlates at over 60 percent with non-competitive tenders for the 28-member European Union (EU). Some of its components are less heavily dependent than others on development, and so government action can modify them more easily, such as reducing red tape or fiscal transparency.
Proxies of IPI can be improved or changed as long as the theoretical model of factors generating opportunities versus constraints remain the baseline, allowing the calculation of a similar index for policy sectors and regions (Drapalova 2017). The index offers a basic starting point for redressing a country’s control of corruption, with the index webpage displaying comparisons on every component for countries from the same income group or region.
While IPI predicts variation across countries successfully and offers a guide for action, does it predict change better than existing indicators? Limitations to the predictive capacity of IPI clearly exist, some by design. First, although all its factors are actionable, some, like e-citizens, are development dependent. Second, as the goal is actionability, IPI overlooks some factors reported in the literature as contributing to corruption, like ethnic fractionalization or natural resources. Due to lack of data other than estimates it also overlooks informality, which is not only a resource for corruption, but may prove a useful reform area, as the example of some recent achievers (like Uruguay) show. Finally, although all components are validated in panel regressions as reported in Mungiu-Pippidi and Dadasov 2016, when running fixed effects it shows that they their predictive value is far greater across countries than across time.
Interpreting IPI change is therefore not very likely to tell us much for the four years interval the index existed. A far better strategy would be to trace its components in time, since we know they are powerful determinants of corruption. Again, difficulties arise due to short historical data on Facebook, and the reduced coverage on budget transparency. To supplant them, we identify two close proxies[2], already tested in Mungiu-Pippidi and Martinez 2014.
– the OSI (Open Service Index) to stand for government transparency, which is an expert assessment of “each country’s national website in the native language, including the national portal, e-services portal and e-participation portal, as well as the websites of the related ministries of education, labour, social services, health, finance and environment as applicable;
– the EPI (e-participation index), which “is derived as a supplementary index to the UN E-Government Survey. It extends the dimension of the Survey by focusing on the use of online services to facilitate provision of information by governments to citizens (“e-information sharing”), interaction with stakeholders (“e-consultation”) and engagement in decision-making processes”.
It includes:
a) E-information: Enabling participation by providing citizens with public information and access to information without or upon demand
B) E-consultation: Engaging citizens in contributions to and deliberation on public policies and services
C) E-decision-making: Empowering citizens through co-design of policy options and coproduction of service components and delivery modalities
Our inquiry strategy therefore is exploratory and meant to offer only some reflection on both indicators and change. On one hand, we associate change in control of corruption and in IPI components, being however aware of their limitations. On the other, we survey continents in detail to select leaders, backsliders and laggards (using criteria involving more than two components) and thus compensate qualitatively for both the lags and what we know are powerful interactions between IPI components as reported in Mungiu-Pippidi and Dadasov, between resources and constraints.
Beyond modernization theory
Lagging and non-actionable indicators are an important problem, but not the only one. Theories of change of governance and of building control of corruption remain vague and poorly proven, plagued by endogeneity or over simplification (Mungiu-Pippidi and Johnston 2018). Modernization theory supporters would expect that governance improves following an improvement in development, proxied by Human Development Index. HDI explains more than half the variance in cross-sectional regressions, with most recent governance achievers and some classic ones as positive outliers (Singapore, Netherlands, New Zealand, Bhutan, Rwanda, Botswana, Chile, Norway), and as negative outliers (performing worse on governance than their HDI would predict Mexico, Argentina, Brazil, Italy, Greece, Qatar, Central Asian Central Asian countries, Russia, China, Vietnam, Myanmar, Chad and Venezuela.
Figure 1. The association between IPI and development
Source: IPI (ERCAS) 2017 and HDI (UNDP) 2016.
The association of IPI with HDI is highly significant, but about 40% of the cases deviate from the rule (Adj Rsq=64). Positive outliers are the classic achievers (Scandinavians, New Zealand), contemporary achievers (Portugal, Costa Rica, Georgia, South Africa) and countries well under good governance still: Ghana, Senegal, Benin, Mali, Liberia and Sierra Leone. The negative outliers cover all development specter, ranging from Greece, Argentina and Qatar to Zimbabwe. The chart uses IPI 2017: in 2019 Myanmar, a negative outlier, has progressed significantly. We interpret this chart as suggestive for the fact that agency is not entirely grounded in development, and can succeed or fail on its own or combined with various contingencies.
We have tested a change in HDI against a change in control of corruption next. The change in HDI fails to predict change of governance as the next three charts (Figure 2-5) illustrate. Two components of HDI tested separately, life expectancy and education show similar poor correlation, despite improving in the time interval studied. In addition, the increase in Internet coverage by itself is not a significant predictor of change in governance.
Figure 2. Change in HDI does not predict change in corruption
Source: World Bank WGI and UNDP.
Legend: association of change in both indicators for ten years (2008-2018), OLS.
Figure 3. Education and control of corruption trends in time
Figure 4. Life expectancy and control of corruption trends in time
Reduction of power discretion
One theory of change grounded in classic corruption scholarship (social choice) argues that corruption might go down if transaction costs are simplified and regulatory quality improved. Improvements in transparency and reduction of administrative burden are supposed to deliver progress. Therefore, we next examine if reductions in transaction costs as proxied by administrative burden and trade openness index, increases in fiscal transparency and e-government took place, and if they predict changes on corruption.
Administrative burden decreased significantly in the last ten years, with major progress by a variety of countries across continents and political regimes: Azerbaijan, Vietnam, Morocco, Tajikistan, China, as well as Costa Rica, Colombia, Brazil, Guatemala, India, Tunisia, Timor-Leste, to name only the top achievers. Giving the nature of the indicator (time and number of procedures to register a business and pay taxes) there is only positive evolution. The least progress is in Europe and Middle East with North Africa, due to very good levels that the regions had already. Latin America (see Figure 5) gained a half point on the 0-10 scale, with Brazil only having still much potential to improve in this area. Progress is also impressive in sub-Saharan Africa, selected Asian countries and particularly Eastern Europe. Sub-Saharan Africa seems to have the greatest association, with older (Rwanda, Ghana) and newer (Senegal, Kenya, Angola, Ivory Coast) progress cases cog the rule that reduction of administrative burden seems to bring about a reduction of corruption.
Figure 5. Reduction of administrative burden across regions
Source: World Bank Doing Business as in Public Integrity Index. Scale 0-10, 10 implying the lowest burden.
When testing improvement of administrative burden versus change in control of corruption the association is significant, but explains very little of the variance. The same happens with trade openness, where there is also quite significant progress in some countries, although less on the whole. When looking at import-export red tape changes over time, we (see Figure 7).
Figure 6. Change in administrative burden versus change in control of corruption
Source: World Bank Doing Business as in Public Integrity Index and World Bank Control of Corruption.
Improvement on red tape might not explain more of Control of Corruption change simply because a part of the change variable that we use (difference across years) is noise resulting from the normalization of original ratings or other problems related to the methodology. Alternative explanations also exist as to the limited capacity of red tape to predict change. Countries game the indicators, without much happening de facto- change remains superficial and barriers simply move elsewhere. North Macedonia, whose application to join EU has just been indefinitely postponed is an illustrative case. The country has reached top of the world in Doing Business World Bank indicator even during the previous regime, which made the object of much ridicule[3]. North Macedonia declined in IPI from 2015 on (7.06 to 6.59), despite being in 2015 the best performer in its income group on Administrative Burden. The reason is what seems to be an unaccountable decline in the two basic political indicators, press freedom (a regional trend) and judicial independence as measured by WEF and used in IPI. This leads to another explanation: Factors that are more powerful offset the administrative burden reduction. South Africa raised on top of the world on fiscal transparency, but registered declines in other areas- similar to Macedonia.
Figure 7. Increase in trade openness across regions
Source: World Bank Doing Business as in Public Integrity Index and World Bank Control of Corruption. Scores are transformed to a 0 – 10 scale, with 10 implying the highest openness.
All its proxies are significant determinants of change, although their explanatory power is small. This is good news, because alongside red tape reduction, e-government (e-services offered) and fiscal transparency, the other proxies for reducing power discretion have also progressed, although less than some expectations. For a continent like sub-Saharan Africa, which started from a very low level, progress on e-government is significant, but has not yet reached the level to make a fundamental difference in the way the state delivers services (see Figure 8). However, when change in online services offered is looked at in relation with the change in control of corruption scores, it explains only very little variation of changes in countries across time.
Figure 8. Global evolution of online services offered
Source: United Nations Open Services Index (2008-2018), scale to 0 – 10, with 10 being the highest score.
Figure 9. Change in online services offered does not correlate with change in corruption
Source: United Nations Open Services Index (2008-2018), transformed to a 0 – 10 scale, with 10 implying the highest score.
There is still much room for improvement on fiscal transparency, even if business registration has progressed. This is the actionable area with the greatest potential, although unfortunately the least documented in this report due to absence of sufficient historical data for Open Budget survey.
Civil society and empowered citizens
We examine next the performance of the Tocqueville based hypothesis, that societies with a higher degree of associativity and collective action capacity have better quality of governance. These are difficult proxies to find, seeing how recent Facebook still is, although a clear connection exists between the rise of global demand for good governance in the streets and social media mobilization. While the growth of Internet does not represent enough empowerment, e-participation as measured by the UN is a proxy for digital citizens’ capacity to exercise demand. It seems to be significant in bivariate regressions and across our continental survey. However, when controlling for development changes its significance is lost. Engaged citizens using e-government tools do seem to improve governance, although the explanatory power of this item (and every other) is small. Over ten years, the global average tripled with changes across all regions, signaling the rise of significant numbers of digitally empowered citizens (see Figure 12). This is a new resource for control of corruption, substituting for classic associations, and needs careful monitoring. So far, if we remove developed countries, numbers are still insufficient for critical mass in many developing nations, but the growth is impressive, including in the poorest continent of sub-Saharan Africa.
While all continents and also the majority of countries evolved to a greater digital citizen empowerment, this alone did not automatically translate to a better control of corruption in countries (Figure 13).
Figure 12. Digital empowerment evolution
Source: United Nations E-participation Index (2008-2018), scale to 0 – 10, with 10 being the highest score.
Figure 13. Digital empowerment and change in corruption
Source: United Nations E-participation Index (2008-2018), scale to 0 – 10, with 10 being the highest score.
Rule of law and judicial independence
The evolution of judiciary independence across a ten year period has not been linear, with an overall trend being slightly negative (Figure 14). Previous literature used for the documentation of the IPI reported an important role of judicial independence, proxied by the World Economic Forum GCR survey item. Countries that have seen an improvement in the independence of the judiciary should thus see an improvement in the Control of Corruption. However, data limitations apply, so testing needs to be restricted to just ten years, and exclude many cases of interest. Nevertheless, the item is significant even with control for development in predicting change in control of corruption, although it explains very little of the variation (Figure 15). For instance, the two countries with most extensive anticorruption campaigns in the world in the last decade, Brazil and Romania, have progressed on Judicial independence (recoded 1-10) about a point each, similar with progress registered by Egypt after the Arab spring and before the assault on the judiciary began, so for a far shorter interval of time and more controversial history.
Figure 14. Evolution of judicial independence 2008-2018
Source: World Economic Forum Juducial Indpendence (2008-2018), scale to 0 – 10, with 10 being the highest score.
Figure 15. Judicial independence evolution vs evolution in control of corruption
Democratic backsliding
There has been much discussion in the latest years about the backslide of democracy. To test if change (negative or positive) in this area can be used to forecast control of corruption we checked primarily freedom of the press, a IPI component, and additionally civil and political rights measured by Freedom House (reported as significant in Mungiu-Pippidi 2015:ch 4). Evolutions in both civil and political rights as assessed by Freedom House are significantly associated with IPI 2019 and change in control of corruption. Negative changes in control of corruption provide a chief explanation of global stagnation of corruption, as freedom of the press has been regressing nearly everywhere for the past ten years, starting with more developed areas, such as Europe and Latin America to the least developed ones. In Europe, Romania only made some progress over the ten years interval, while Hungary, Greece, Bulgaria and Poland regressed. In Africa, there is more balance between negative and positive change, but seeing the dismal departure level of change the situation does not improve very much. MENA and Eastern Europe also do not do well (averages in the negative and little to no positive change), and freedom of the press offsets China or Vietnam’s advancement on other indicators, although they both progress on IPI since 2015. Figure 16 shows a decline over all continents.
When we correlate the change in Freedom of the Press with the change in control of corruption (Figure 17), the relationship shows as significant, albeit explaining around 14% of the change across time.
Figure 16. Freedom of the press involution across continents
Source: Freedom House Freedom of the Press (2007-2017), transformed to a 0 – 10 scale, with 10 implying the highest freedom.
Figure 17. Freedom of the Press involution vs change in control of corruption
5. Can we trace change in corruption using perception indicators?
The world of lagging control of corruption and lagging corruption indicators looks like the two charts below, which average change across income groups and continents. The world seems entirely flat, but for the two higher income groups where the trendline is declining. However, this is not a significant decline, indicating that the world is really flat when governance is concerned.
1. Control of Corruption World Governance Indicator
Ten Year Trends by Income Group
2. Control of Corruption World Governance Indicator
Ten Year Trend by Continent
During the period represented by this relatively static trend line, the world has invested more than ever in good governance following the adoption of United Nations Convention against Corruption in 2004. While we do not have country success stories in abundance, it is also implausible that no evolution at all has taken place. We need, however, better tools to trace this evolution
Why not use the Index for Public Integrity, created by ERCAS in 2016? Well, for the very simple reason that the IPI is still relatively recent – some of its components do not go far back enough to allow us to compile it retrospectively. Also, despite its superior specificity (we know exactly what components it includes, what do each measure and how they interact), the IPI is also an aggregate, and the components are also standardized and normalized to allow a ranking in any given year, thus generating statistical noise itself, even if less than the perception indicators.
However, it remains very important to trace the evolution of control of corruption in time. Various government agencies use governance indicators to program and condition foreign aid. Governments and civil societies also need a tool to gauge the effectiveness of their policies. How can we reliably have an instrument that both captures change more sensitively and tells us what drives the change?
6. How can a more specific and sensitive forecast be developed?
To solve this riddle, we proceed as follows:
1. TREND ANALYSIS AND THRESHOLD TEST
We use the disaggregated components of IPI, and we observe their changes since 2008. To eliminate changes which may just be random we compare our sample of 120 countries with a similar theoretical group of countries where average change is zero (like in the graphs above). We rate as significant change any change above or below the global standard deviation of average change against a control group with zero change. This is a more positive scenario than using a null hypothesis with average global change as baseline, but still eliminates small changes. It is also less arbitrary than just setting a confidence interval above/below which we would consider changes too small. As this exercise takes place in 2021 amidst a context of global democratic decline we opt for this variant because our indicators have mixed signs: some have negative average changes (political indicators), others positive (more technical indicators, like e-citizens). We do not want to miss reforms, but to encourage countries to engage in them by making them visible. As a complement to the Country page trends table where this approach is used, you can compare each country against the regional (continental) mean by using this Compare Trends button on the forecast map page.
We then rate change as consistent if a country has progressed (or regressed) in at least two indicators and had not regressed (or progressed) in any. Some data is missing retrospectively: the Facebook users’ data, which is a component of the e-citizens indicator changes quality and coverage across years, so we use only Internet household connections to measure e-citizens; also, the World Bank has repeatedly changed methodology for Trade Openness, so we do cannot use this IPI indicator for the trend analysis.
2. POLITICAL CHANGE CHECK
In case of inconsistency, as a mandatory step we check all radical political events (regime change or political violence) which occurred in the last 4 years in a country and if the sign of these events is in line with or contradictory to the long-term trends. A plus on the long-term trends can be canceled by a minus on political change, for instance a coup, a violent repression of civil society. Likewise, a negative on the long-term trends can be counter-balanced by an uprising that is followed by the election of an anticorruption government.
3. SOCIETAL DEMAND (NORMATIVE CONSTRAINTS) CHECK
Finally, evidence shows that the demand for change from the society matters a lot. In case of inconsistencies, we thus use as final check the most recent value for the IPI e-citizens (digital citizens) component as proxy for demand for good governance. Additionally, for particularly puzzling cases or where there is change, but it is under the statistical threshold, we use the latest edition of Transparency International’s Global Corruption Barometer, which includes a question where citizens are asked if they perceive a change, and if so in what direction. They are also asked if they approve of the anticorruption stance of the government – all useful questions to estimate demand for change in a society.
What are the results?
The trend analysis and weighting described above produces three categories of countries: stationary cases, improvers and backsliders. These can be further refined by plotting the change sign against the IPI map, so to see if stagnation, for instance, occurs where such lack of change is problematic, such as where there is already systemic corruption. It is not a problem if a country with good public integrity stagnates.The forecasted trends are indicated as either red (declining) or green (improving). You can access the IPI map on this webpage under Forecast. The final categories are established by our senior experts, professors Alina Mungiu-Pippidi and Michael Johnston, and reviewed by a group of experts knowledgeable in all continents and governance indicators.
This exercise returns the results that we hoped for. We identify change in over 30 countries, and encouragingly in over 20 of them this change is positive. For some, like Japan or Estonia, corruption is already rather the exception than the rule, while others have far more to travel to good governance, but they are on a positive trend. For a detailed legend of why a country changed or did not, please consult the country pages under Forecast.
Improvers and decliners in the 2022 forecast
| Improving | Declining |
| Bulgaria, Costa Rica, Estonia, Latvia, Lithuania, Indonesia, Japan, Kenya, R. Korea, Kyrgyz Republic, Liberia, Moldova, Mongolia, Morocco, North Macedonia, Slovakia, Spain, Uruguay, Taiwan, Timor Leste, Vietnam, Zimbabwe | Bolivia, BiH, Cambodia, Egypt, Ethiopia, Myanmar, Russian Federation, Venezuela, Zambia |
7. What to use this forecast for?
The forecast can serve as an evaluation tool for the anticorruption strategists in a country, as well as for a longer-term diagnosis completing the Index for Public Integrity, which offers only a snapshot in one moment in time. It is important to understand the trend a country is on to confirm or adjust your theory of change and your strategy accordingly. For some very basic choices for donors and civil societies, see the table below.
©Alina Mungiu-Pippidi and ERCAS. All rights reserved
[1] Available at www.integrity-index.org
[2] https://publicadministration.un.org/egovkb/Portals/egovkb/Documents/un/2016-Survey/Annexes.pdf
[3] https://www.esiweb.org/index.php?lang=en&id=67&newsletter_ID=85
References
Andersson, S., & Heywood, P. M. (2009). The politics of perception: use and abuse of Transparency International’s approach to measuring corruption. Political Studies, 57(4), 746–67.
Gächter, S., & Schulz, J. F. (2016). Intrinsic honesty and the prevalence of rule violations across societies. Nature, 531(7595), 496.
Johnston, M. (2010). Assessing vulnerabilities to corruption: Indicators and benchmarks of government performance. Public Integrity, 12(2), 125–42.
Kraay, A., & Kaufmann, D. (2002). Growth without governance. The World Bank. Available here
Kaufmann, D., Kraay, A., & Zoido-Lobatón, P. (1999). Aggregating governance indicators (Policy Research Working Paper 2195). Washington, DC: The World Bank.
Kaufmann, D., Kraay, A., & Mastruzzi, M. (2007). Measuring corruption: Myths and realities (Africa Region Findings and Good Practice Infobriefs 273). Washington, DC: The World Bank.
Mungiu-Pippidi, A. (2015b). The Quest for Good Governance: How Societies Develop Control of Corruption. Cambridge, UK: Cambridge University Press.
Mungiu-Pippidi, A., & Dadašov, R. (2016). Measuring control of corruption by a new index of public integrity. European Journal on Criminal Policy and Research, 22(3), 415–38.
Mungiu-Pippidi, A., & Dadašov, R. (2017). When do anticorruption laws matter? The evidence on public integrity enabling contexts. Crime, Law and Social Change, 68(4), 387–402.
Mungiu-Pippidi, A., & Johnston, M. (eds). (2017). Transitions to Good Governance: Creating Virtuous Circles of Anti-corruption. Cheltenham, UK: Edward Elgar Publishing.
UNDP [United Nations Development Programme]. (2008). A users’ guide to measuring corruption. Oslo: UNDP Oslo Governance Centre.
Treisman, D. (2007). What have we learned about the causes of corruption from ten years of cross-national empirical research? Annual Review of Political Science, 10, 211–44.
Appendix
