|
|
|
|
|
This page provides answers to the most relevant and most commonly raised issues about the Worldwide Governance Indicators contained in the series papers:
|
|
| Answers | |
|
Governance can be broadly defined as the set of traditions and institutions by which authority in a country is exercised. This includes (1) the process by which governments are selected, monitored and replaced, (2) the capacity of the government to effectively formulate and implement sound policies, and (3) the respect of citizens and the state for the institutions that govern economic and social interactions among them. For more information, consult Appendix D of the Governance Matters VIII paper.
|
|
What are the 6 dimensions of the Worldwide Governance Indicators? The six dimensions of Governance are: Voice and Accountability; Political Stability and Absence of Violence; Government Effectiveness; Regulatory Quality; Rule of Law; and Control of Corruption. Voice and Accountability measures the extent to which country’s citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media Political Stability and Absence of Violence/Terrorism measures the perceptions of the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including domestic violence and terrorism Government Effectiveness measures the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies Regulatory Quality measures the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development Rule of Law measures the extent to which agents have confidence in and abide by the rules of society, in particular the quality of contract enforcement, the police, and the courts, as well as the likelihood of crime and violence Finally, Control of Corruption measures the extent to which public power is exercised for private gain, including petty and grand forms of corruption, as well as “capture” of the state by elites and private interests. For further details, consult Appendix D of the Governance Matters VIII paper.
|
|
How frequently are the Worldwide Governance Indicators updated? The Worldwide Governance Indicators were updated every two years between 1996 and 2002. After 2002, they are updated on a yearly basis. All relevant information (including data, methodological papers, interactive charts, and world maps) for the last round of updates (2009) is now posted on the web at: www.govindicators.org. The next round of Worldwide Governance Indicators will be posted in mid-2010.
|
|
Will you go further back in time to compile Worldwide Governance Indicators for years previous to 1996? No, 1996 will remain our starting year. As we go back in time, we would have to drop several of our sources as they became available only in recent years. Dropping sources would decrease both the precision of our estimates (i.e. higher standard error) and the interpretation of changes over time (as a country relative position could be affected by the subtraction of sources rather than an actual change in its performance).
|
|
Is data from the underlying sources used to compile the indicators available? Yes, beginning from the 2006 release, we are making public all the sources that were used to aggregate our indicators. To access this data, visit our interactive website. In the output page view the statistical table and click on the source link. An expanded table will be reporting information about all the selected sources, including the underlying ratings.
|
|
Why are margins of error important? Inherent to all Governance Indicators is a margin of error, which might vary from country to country, normally attributable to two factors: (i) cross-country differences in the number of sources in which a country appears, and (ii) differences in the precision of the sources in which each country appears. In spite of the considerable number of individual sources used (which tends to decrease the extent of measurement error), there are still substantial margins of error associated with governance estimates. This implies among other things that it is difficult to assign many countries to a definitive performance category according to their estimated level of governance, and even more difficult to compile precise rankings. It should be emphasized however that over time the standard errors have been sensibly reduced thanks to the increase number of sources utilized. Indeed, while average standard errors in 1996 averaged 0.34 across the 6 Indicators, in 2008 the figure was reduced to 0.21. It is also very important to notice that the margins of error we emphasize are not unique to the perceptions data we use to construct our aggregate governance indicators: measurement error is pervasive among all measures of governance and institutional quality. An advantage of our measures of governance is that we are able to be explicit about the accompanying margins of error, whereas these are most often left implicit with objective measures of governance. For a more thorough discussion, consult the Governance Matters III paper (pages 12-15) and the Governance Matters IV paper (pages 27-31).
|
|
Is the precision of these Worldwide Governance Indicators higher than for others, and is it improving over time? Perceptions-based or subjective measures of governance contain important information often not captured by objective indicators, particularly in emerging economies. For example, we show in the Governance Matters IV paper that the firm’s perceptions of the difficulty of starting a new business, or of their tax burdens, do not depend solely on the relevant legal framework governing business entry and taxation. Rather, firms’ views on these issues are also importantly influenced by the degree of corruption in their country (particularly so in developing countries), suggesting that not only do formal rules matter, but also the institutional environment in which these rules are applied and enforced. As a result, it should not surprise that the precision of the Worldwide Governance Indicators that rely on subjective data is higher than other objective Indicators. In the Governance Matters III paper, we designed simple exercises to show that objective Indicators could indeed be exposed to margins of errors much larger than those constructed on the basis of subjective data. It should also be emphasized that over time the standard errors have been sensibly reduced thanks to the increase number of sources utilized. Indeed, while average standard errors in 1996 averaged 0.34 across the 6 Indicators, in 2005 the figure was reduced to 0.21. For more details, consult the Governance Matters IV paper (pages 27-31) and the Governance Matters III paper (pages 12-15).
|
|
Why do you use subjective measures as opposed to objective indicators? The primary reason for this choice is that for many of the key dimensions of governance, such as corruption or the confidence that property rights are protected, objective data are almost by definition impossible to obtain, and so there are few alternatives to the subjective data on which we rely. Perceptions-based or subjective measures of governance contain important information often not captured by objective indicators, particularly in emerging economies. For example, we show in the paper that the firm’s perceptions of the difficulty of starting a new business, or of their tax burdens, do not depend solely on the relevant legal framework governing business entry and taxation. Rather, firms’ views on these issues are also importantly influenced by the degree of corruption in their country (particularly so in developing countries), suggesting that not only do formal rules matter, but also the institutional environment in which these rules are applied and enforced. For more details, consult the Governance Matters IV paper (pages 27-31) and the Governance Matters III paper (pages 22-24).
|
|
How do I interpret changes in countries' estimates over time? A change over time could be attributed to 4 factors. First of all, it could come from a change in the score assigned to a country by the same source over time. This is the most common and relevant factor, directly reflecting changes in perceptions of the country's performance. A second factor is the addition of new sources whose ratings might be different from the average ratings from pre-existing sources. Changes over time in relative performance may also reflect the addition of new countries to the aggregate indicator. If for example we add a country with a governance rating that is high relative to those countries already in the index, then by construction all the countries which rank lower than this country will receive lower scores. Finally, a change in a country's performance could derive from a change in the weights in the aggregation procedure. Overall, however, these last two factors typically have only very small effects on changes. The first factor is by far the most relevant. To show this, for each country we have summarized the extent to which changes in the individual sources agree with the direction of change in the aggregate indicator by building an 'agreement ratio', calculated as the number of sources that agree with direction of change divided by the sum of the number of sources that agree and number of sources that disagree with direction of change. We find that the agreement ratio is quite high for countries with large changes in governance. Averaging across all countries and indicators, we find an average agreement ratio of 0.86 for the period 1996-2005. For the six indicators, the agreement ratio ranges from a low of 0.76 for Government Effectiveness to a high of 0.93 for Voice and Accountability. This provides some confidence that for countries with large changes in our governance estimates, these changes are being driven primarily by changes in underlying sources. For more details, consult the Governance Matters III paper (pages 16-19) and the Governance Matters IV paper (pages 10-13).
|
|
How confident can we be that over time changes are indeed significant? The margins of error associated with levels of governance are substantial. Since changes over time are in most cases small relative to levels of governance, it is safe to assume that most of the observed changes over time are neither statistically nor practically significant. However, there are some cases where the changes over time are large enough that the 90% confidence intervals in the two periods do not overlap. This rule of thumb helps to identify cases of changes over time that are likely to be of practical significance. We develop a formal statistical methodology, as well as some simple rules of thumb, for identifying changes in governance that are likely to be statistically and practically significant. Over the eight-year period spanned by our governance indicators, we find that in about 10 percent of countries we can be highly confident (at the 90 percent significance level) that governance has changed substantially, while at a lower 75 percent significance level, roughly 20 percent of all observed changes stand out as significant. Importantly, we show that there is a great deal of agreement among our many data sources about the direction of change in governance in these countries. Overall this reminds us that, while in general institutional quality changes but gradually, there are also countries where one can point to sharp improvements or deteriorations over an eight-year period. This finding is of particular interest given the common perception that, while deterioration in a particular country can take place rather quickly, improvements are always very slow and incremental. A more detailed discussion of confidence intervals, standard errors and changes over time can be found in the Governance Matters III paper (pages 14-19, see also figure 3 on pages 49-51) and the Governance Matters IV paper (pages 15-26).
|
|
What are the underlying sources for the Worldwide Governance Indicators? Our data sources reflect the perceptions of a very diverse group of respondents. For 2009, we use 440 variables drawn from 35 sources and 33 different organizations. Several of our data sources are surveys of individuals or domestic firms with first-hand knowledge of the governance situation in the country. We also capture the perceptions of country analysts at the major multilateral development agencies, reflecting these individuals’ in-depth experience working on the countries they assess. Other data sources from NGOs, as well as commercial risk rating agencies, base their assessments on a global network of correspondents typically living in the country they are rating. For further details, consult the Appendices of the Governance Matters VIII paper.
|
|
| In simple terms, how is the aggregation methodology carried out to produce Governance estimates?
We use an Unobserved Component Model (UCM) to aggregate the various response in the broad 6 clusters. This model treats the "true" level of governance in each country as unobserved, and assumes that each of the available sources for a country provide noisy "signals" of the level of governance. The UCM then constructs a weighted average of the sources for each country as the best estimate of governance for that country. The weights are proportional to the reliability of each source. The resulting estimates of governance have an expected value (across countries) of zero, and a standard deviation (across countries) of one. This implies that virtually all scores lie between -2.5 and 2.5, with higher scores corresponding to better outcomes. For technical details, consult Appendix D of the Governance Matters VIII paper.
|
|
|
What is the impact on your aggregate indicators of some underlying sources being less precise than others?
The Unobserved Component Model (UCM) that we use to aggregate the various response in the broad 6 clusters constructs a weighted average of the sources for each country, where the weights are proportional to the reliability of each source. Therefore, the model minimizes the margins of error by automatically assigning lower weights to those sources that have larger noise and/or measurement errors. For technical details, consult Appendix D of the Governance Matters VIII paper.
|
|
How many countries are covered by the Worldwide Governance Indicators? Coverage varies depending on the indicator and the year. For 2008, Government Effectiveness has the largest coverage (212 countries), while Regulatory Quality and Control of Corruption have the smallest (208). For a complete list of countries for each Governance Indicator and their governance ratings, consult Appendix C of the Governance Matters VIII paper.
|
|
| Can we infer any global trend over time from the Worldwide Governance Indicators? Our indicators measure governance in units where the average score for the world as a whole is zero in every period. Therefore, the Worldwide Governance Indicators are only meant to capture countries' relative position vis-a-vis the others. We can however use our indicators jointly with our underlying sources to draw conclusions on broad global trends. For instance, since many of our individual sources show a deterioration over time in worldwide averages, then we can safely infer that a country's deterioration in its relative position cannot be attributed to an overall improvement in other countries, but rather is likely to reflect a poorer performance by the country. For a more detailed discussion, consult the Governance Matters III paper (pages 8-12) and the Governance Matters IV paper (pages 13-14).
|
|
| How does the introduction of 'persistence' affect the interpretation of changes over time?
There are two types of persistence which tend to have opposite effects upon the significance of any observed change in data: persistence in governance and persistence in the measurement error. Persistence in governance is quite common. Quality of institutions tend to change very slowly. Therefore the governance frameowrk in any given country tends to be highly correlated with previous levels. The introduction of persistence in governance, however produces large effects upon the interpretation of significance of changes. Given any observed change in governance levels, the higher the persistence in governance, the more likely that any such change is the result of pure noise and therefore less likely to signal a significant change in unobserved governance. In the limit where governance is perfectly correlated in the two periods, we would know for sure that any change observed in the data must reflect only fluctuations in the error term, and so we would completely discount the observed change in the data. On the other hand, persistence in the error term can produce symmetrically opposite results. The emergence of persistence could occur, for example, in the presence of methodological flaws in some of the individual sources use to measure the governance score. Given any observed change in governance levels, the higher the persistence in the error term, the more likely that any such change understates the true change in unobserved governance. Overall, we find that the effect of persistence in governance tends to dominate the other, therefore adding one further dimension of caution in interpreting the significance of changes over time. For a more detailed discussion, consult the Governance Matters IV paper (pages 16-20).
|
|
Why do you distinguish between representative and non-representative sources? This distinction allows for minimization of the imprecision of point estimates due to measurement errors in underlying sources. First of all, non-representative sources are more likely to be subject to higher measurement error given their more limited scope (for instance, a source rating only rich countries might give them lower ratings than other sources covering a more balanced panel of low and high-income countries). On a more technical note, the distribution of unobserved governance in the subset of countries covered by these smaller-scope sources is not the same as that in the world as a whole. As a result, for these sources we cannot make the assumption that unobserved governance in the countries covered by these surveys follows a standard normal distribution, as is required by the maximum likelihood procedure. For a more detailed discussion, consult the Governance Matters III paper (pages 8-12).
|
|
How can I see what variables were actually used to compile the latest set of Worldwide Governance Indicators? Appendix A of the Governance Matters VIII paper lists all the sources that were used, along with a brief description and weblink to the respective homepages. Appendix B of theGovernance Matters VIII paper instead provides details on how we have assigned individual questions from each of these sources to our six governance clusters.
|
|
| What options do I have to access the data?
Appendix C of the Governance Matters VIII paper provides a printout of all the data. For each indicator, it shows for each country the estimate level, the standard error and the number of sources used in each year. Alternatively, you can download the complete dataset in excel format directly from this website or only specific data tailored to your needs. To do so, click here. For a more complete access to data, charts and related information, visit the interactive website.
|
|
What implications can we draw in regards to the Millennium Challenge Account (MCA)? The MCA allocation rule is designed to ensure that MCA funds will be allocated to low-income countries with relatively sound policies and institutions. A group of 70 countries that are eligible for concessional IDA lending from the World Bank, and which have per capita incomes less than $1435 in 2005, will potentially be eligible for MCA funds in 2005. According to the MCA eligibility rules, this set of countries will be rated according to 16 performance criteria covering three dimensions of performance: "governing justly" (6 criteria), "investing in people" (4 criteria), and "promoting economic freedom" (6 criteria). Four of the Governance Indicators we have constructed (voice and accountability, government effectiveness, rule of law, and corruption) have been proposed as performance indicators under the MCA's "governing justly" performance dimension, with the remaining two for this dimension being measures of civil liberties and political rights constructed by Freedom House. In addition, a fifth governance indicator, Regulatory Quality, is included under "promoting economic freedom". In order to qualify for MCA assistance, countries must (a) be in the top half of all potentially eligible countries according to the corruption rating from the governance indicators, and (b) must be in the top half of all potentially eligible countries on at least half of each of the performance criteria under each of the three dimensions of performance. This rule is designed to ensure that resources are channeled towards countries that are performing well in a variety of dimensions of governance, and in which corruption especially is relatively low. However, it is important to note that the substantial margins of error associated with governance estimates mean that it is difficult to assign many countries to a definitive performance category according to their estimated level of governance. This point applies to any of the MCA criteria. Given the presence of non-trivial margins of errors, some countries ranked under the median might in fact belong to the top half of the distribution. Classifications based on individual indicators, or even on a single aggregate indicator, inevitably run the risk of mis-classifying countries due to the margins of error inherent in all indicators. This underscores the need for a certain degree of flexibility in the MCA allocation rule, in light of the importance of caution when using governance indicators to classify countries into groups. To reduce the risk of misclassification, it is important to look at a variety of indicators and additional sources of data, especially for borderline cases. As an illustration, consider the Control of Corruption indicator. The US government’s Millennium Challenge Account aid program requires recipient countries to score above the median of a group of 70 potentially-eligible countries on this indicator. We can use our estimates of governance and their margins of error to assess the likelihood that corruption in a country actually falls above the median or not. Using our 2005 data, we can identify a group of 17 poorly-performing countries, or about one-quarter of the sample, where there is less than a 10 percent chance that corruption in these countries actually falls above the median. For another 23 countries, or about a third of the sample, we are quite confident that corruption in these countries falls above the median, with a probability of at least 90 percent. In contrast, for the remaining 30 countries, the probability that they fall above the median is somewhere between 10 percent and 90 percent, and so we have less confidence that these countries are correctly classified. If we relax our standards of significance to 25 percent and 75 percent, we find that only about 20 countries out of 70, or 29 percent of countries fall in this zone of uncertainty. This example shows that we can use this kind of data to identify with considerable confidence groups of strong and weak performers. But at the same time the presence of margins of error reminds us that finer distinctions among countries near the middle of the pack are much more difficult to make given the inherent difficulties of measuring governance with any type of data. Fortunately, the decrease of margins of errors over time (due to the increase in sources) have reduced the number of countries with a non trivial probability of mis-classification as explained in the Governance Matters IV paper (pages 8-10). We first performed these MCA-related calculations in late 2002, shortly after the announcement of the initial MCA eligibility criteria. At that time, using the older version of our 2000 Control of Corruption indicator, we found that 23 out of 61 countries (or 38 percent of countries) fell in this intermediate zone. This much higher proportion of intermediate countries reflected the fact that the old version of or 2000 Control of Corruption indicator relied on substantially fewer data sources than we now have available to us for both 2000 and 2005. For a more detailed discussion, consult also the Governance Matters III paper (pages 26-29).
|
|
How have potential ideological biases in poll agencies' ratings been addressed? We address this issue as follows. Our identifying assumption is that surveys of firms or individuals are not tainted by ideology, since they reflect the views of a large number of respondents in each country. In contrast, it is possible that the views of a smaller number of raters affiliated with a particular institution may reflect the ideology of that group. We can therefore identify the effects of ideology by looking at the correlation across countries between the ideology of the government in power, and the difference in the percentile ranks assigned to countries by a poll of experts and a survey of individuals and firms. We implement this idea using the World Bank's Business Environment Survey for 2000, and an indicator variable that takes on the value 1 if the government in power is left-of-center, 2 if it is center, and 3 if it is right-of-center, taken from the database of political institutions constructed by Beck et. al. (2001). The coefficient on the ideology variable will therefore capture the extent to which a given poll of experts rates countries countries with left- or right-wing governments systematically differently from a survey (a positive coefficient indicates that the poll in question tends to rate right-of-center governments more highly relative to a survey). Our results showed that we find only one source which appears to have a consistent ideological bias, with the Heritage Foundation assigning relatively higher scores to countries with right-of-center governments than the corresponding survey and this "ideology bias" is fairly modest in magnitude. For a more detailed discussion, consult the Governance Matters III paper (pages 24-25).
|
|
What is the criteria to assign different colors to countries in the interactive charts and worldmaps? Each country color pattern follows a simple quartile distribution (for illustrative purposes): the best quartile (over 75th percentile) is in green (with top 10% colored in darker green), the second best quartile (over 50th) is in yellow, the third (over 25th) is in orange, and the fourth is in red (with bottom 10th in darker red). Please note that this simple color coding does not account for the size of the confidence intervals; the color mapping is based on the point estimates. To access interactive charts and/or maps, visit our interactive webtool at www.govindicators.org.
|
|
How confident can we be that rankings drawn from point estimates are accurate? Because of margins of errors, we cannot make precise rankings. However, we can still make inferences based on confidence intervals. If we for instance divide the distribution of countries' estimates into two categories (low-high governance rating), we can calculate the probability that any given country might indeed belong to the opposite side of the distribution. A more detailed discussion of confidence intervals, standard errors and rankings can be found in the Governance Matters III paper (pages 11-14, see also figure 1 on pages 45-47) and the Governance Matters VIII paper.
|
|
How confident can you be that sources are independent from each other, as assumed in your aggregation process? It is true that an important assumption of our Unobserved Component Model is that the errors are independent across sources. This assumption in particular imposes the identifying assumption that the only reason why two sources might be correlated with each other is because they are both measuring the same underlying unobserved governance dimension. We have taken several steps to ensure that this assumption would hold. In the first place, we have avoided including sources which were themselves constructed upon other indicators used in the aggregation procedure. For instance, we did not include the Corruption Perceptions Index (CPI) by Transparency International because the CPI is itself an aggregate of a number of individual sources, all of which were included separately in our corruption indicator. Secondly, we were very cautious in flagging risk rating agencies who would base their own assessments on the assessments of other agencies included in our sample. We have to the best of our knowledge excluded any source of governance data where we found it was explicitly based on another one of our sources. If errors are positively correlated across sources despite the precautions we have taken, it would imply that the margins of error we have constructed are conservative, and that the true level of imprecision of the indicators could be larger than we have estimated. However, we have shown in the 2006 paper that there is at best weak evidence of correlated perceptions errors and that even with correlated errors each sources does carry information which should be usefully added to our indicators. For more details, consult the Governance Matters V paper (pages 19-31).
|
|
| What is the best use I can make of these indicators?
Notwithstanding the substantial increase in data collection since 2002, which has both expanded country coverage and improved the precision of the aggregate indicators, margins of error remain. We hope that in the future the availability of additional data will enable further improvements in precision. However, the presence of margins of errors imply that we cannot make precise rankings of the countries solely based on the point estimates. The Worldwide Governance Indicators however can serve the purpose of providing individual countries with a set of monitorable indicators of governance they can use to benchmark themselves against other countries and over time. We recognize there are limitations to what can be achieved with this kind of cross-country, highly-aggregated data. Therefore, this type of data cannot substitute for in-depth, country-specific governance diagnostics as a basis for policy advice to improve governance in a particular country, but should rather be viewed as a complementing tool. A more detailed discussion of confidence intervals, standard errors and rankings can be found in the Governance Matters III paper (pages 11-14 and pages 45-47) and the Governance Matters VIII paper.
|
|
It may be useful to measure it, but does governance really matter? It matters enormously. We find that a country improving its quality of governance from a low level to an average level can in the long term quadruple the income per capita of its population, and similarly reduce infant mortality and illiteracy. And the direction of causality is clear: it goes from etter governance to higher incomes, and not vice versa. In other words, governance is not a 'luxury' good that only wealthier countries can afford; is not the automatic result of development. To the contrary, it requires continuous political will and commitment, and difficult work. There is by now a strong consensus among both academics and policymakers that good governance provides the fundamental basis for economic development. Academic research has focused on the effects of institutional quality on growth in the very long run, noting that there is a strong causal impact of institutional quality on per capita incomes worldwide. Estimates of this “development dividend” of good governance suggest that a realistic one-standard-deviation improvement in governance would raise incomes in the long run by about two- to three-fold. Such an improvement in governance by one standard deviation is feasible and realistic, since it is only a fraction of the difference between the worst and best performers, and would correspond, for instance, to an improvement in the current ratings of Voice and Accountability between the level of Myanmar to that of Kazakhstan, or from the level of Kazakhstan to that of Indonesia, or from the level of Indonesia to that of South Africa. For improvements in Rule of Law, a one standard deviation difference would for instance constitute the improvement from the levels of Somalia to those of Nigeria, or from Nigeria to Macedonia, or from that of Macedonia to Greece, or from Greece to Canada. Even over much shorter periods such as the past 10 years, countries with better institutional quality have grown faster. Of course, there is variation around these relationships, since governance is not the only thing that matters for development – but it certainly is a very important factor deserving policymakers’ attention. For a more detailed discussion, consult the papers Governance Matters VIII, Governance Matters IV, Governance Matters and Growth Without Governance.
|
|
Why are there a few countries with ratings above 2.5 or below -2.5? Given our assumption about governance being normally distributed, there is a 99% chance that a country's rating would fall between -2.5 and 2.5. However, under very extreme circumstances, a country's rating might exceed these thresholds. This simply means that the country has an extremely poor record (below -2.5) or extremely good record (above 2.5) in that specific governance indicator.
|
|
Should weak governance performance in poor countries be discounted because of low income levels? In recent years the international community has rightly turned its attention to the problems of underdevelopment in Africa. Not only is Africa poorer than other regions in the developing world, it also lags starkly behind other regions in terms of progress towards the Millennium Development Goals. If past trends continue, many countries in Africa will need to double their per capita incomes over the next decade in order to attain the goal of halving poverty by 2005. There is widespread consensus that a combination of substantial aid inflows, together with concerted domestic policy effort, is necessary to meet this challenge. In light of the strong positive effect of governance on development, and in light of its importance for effective aid delivery, it is then a matter of considerable concern that governance performance in Sub-Saharan Africa is on average quite weak. Countries in Africa are poor, and too often they are also poorly governed. 38 out of 46 countries in the region are shown to be both poorer than the world average and also exhibit worse governance than the world average. Some observers have argued that we should discount the poor governance performance of the region based on the fact that these countries have very low income levels – arguing that good governance costs money to provide. Yet, as described in the Governance Matters IV paper (pages 36-38), recent research provides very little evidence in support of the proposition that poor governance in Africa is attributable to Africa’s poverty. Rather, most of the action is in the opposite direction, from better governance to better development outcomes.
|
|
What are the 'halo effects' and what impact do they have upon the strong positive correlation between governance and income? Perceptions-based measures of governance such as the ones we develop are potentially subject to a number of biases. One common critique is that perceptions of governance are biased upwards in rich countries because respondents view the development success of the country in question as evidence that institutional quality is good. This type of bias is sometimes referred to as a “halo effect”, and some observers have argued that such “halo effects” might be significantly responsible for the highly positive correlation between income and governance Yet, as described in the Governance Matters IV paper (pages 32-36), we argue that such halo effects would need to be implausibly large to account for cross-country correlations between governance and incomes.
|
|
Why do some countries get low scores on Political Stability and Absence of Violence/Terrorism even though they appear to be quite stable? This indicator does not measure how long a particular government has been in power. Instead, it captures perceptions of the likelihood of politically-motivated violence, including terrorism. Thus the United States for example has a sharp decline in this dimension between 2000 and 2002. This happened not because the political process in the United States is now perceived as more unstable than in the 1990s. Rather, it reflects perceptions of the risk of terrorist attacks on the United States that increased sharply following the events of September 11, 2001. Similarly, countries that are functioning democracies, but are marred by domestic politically-motivated violence, may also not score well on this indicator.
|
|
How can some countries register large declines in their rankings, but these declines are not statistically significant? When comparing country scores over time, or with other countries, due regard must be paid to margins of error. A simple rule of thumb is that when the indicated confidence intervals do not overlap for the two comparators, the change is not statistically significant. For some countries, particularly where data is scarce and our estimates are based on very few data sources, these confidence intervals can be quite large, properly indicating the uncertainty about governance in these countries. For such countries, even apparently large changes may not be statistically significant, simply because there is a great deal of transparently-noted imprecision in the data. A more detailed discussion of confidence intervals, standard errors and changes over time can be found in the Governance Matters III paper (pages 14-19, see also figure 3 on pages 49-51) and the Governance Matters IV paper (pages 15-26).
|
|
How do the Worldwide Governance Indicators relate to the World Bank's new Governance and Anticorruption Strategy? On March 21, 2007 the Board of the World Bank approved a new Governance and Anticorruption Strategy (GAC) to help countries and aid donors tackle governance challenges worldwide. Bank staff are currently developing an implementation plan. See www.worldbank.org/wbi/governance for details. The WGI have no official status in the World Bank in general or with the GAC in particular. The WGI are one of a large number of empirical tools that can be used to monitor governance, in the context of the GAC as well as in many other contexts.
|
|
 © 2009 The World Bank Group, All Rights Reserved. Legal |
Designed and developed by WBI
Multimedia Center |
