O- A -- POLICY RESEARCH WORKING PAPER 2772 Governance Matters II updatedg,;ernari e indicators report eJrirnlaLtes of six dimensions of governance Updated Indicators for 2000-01 for 175 countries In C'01-D-0 I They can be compared with Daniel Kaufmann those constructed for Aart Kraay 1997-98. Pablo Zoido-Lobat6n The World Bank Development Research Group and World Bank Institute Governance, Regulation, and Finance Division February 2002 POLIcy RESEARCH WORKING PAPER 2772 Summary findings Kaufmann, Kraay, and Zoido-Lobat6n construct 2196, October 1999) to newly available data to arrive at aggregate governance indicators for six dimensions of governance indicators comparable with those governance, covering 175 countries in 2000-01. They constructed for 1997-98. The data is presented in the apply the methodology developed in Kaufmann, Kraay, appendix, and accessible through an interactive Web- and Zoido-Lobat6n ("Aggregating Governance interface at http://www.worldbank.org/wbi/governance/ Indicators," Policy Research Working Paper 2195, and govdata2001.htm. "Governance Matters," Policy Research Working Paper - - paper-a joint product of the Development Research Group and the Governance, Regulation, and Finance Division, World Bank Ilnstitute-is part of a larger effort in the Bank to develop and analyze governance research indicators and trends worldwide. Copies of the paper are available free from theWorld Bank, 1818 H StreetNW, Washington, DC 20433. Please contact Erin Farnand, room J3 -280, telephone 202-473-9291, fax 202-334-8350, email address efarnandCaworldbank.org. Y olicy Research Working Papers are also posted on the Web at http://econ.worldbank.org. For access to the data and related napers, visir httpv . worldbank.org/wbi/governance/pubs/govmatters200l.htm. The authors may be contacted at ., ,I,,; l;;".o. Il-.. ., or akraay@uworldbank.org. February 2002. (53 pages) The To ICy Resea1-rch Paper Series disseminates the - of work in progress to encourage the exchange of ideas about deve/iopmo;.a issues .Lln oib7jective of the series is tog et thefindings outquickly, even if the presentations are less than fully polished. The p)apers .orri t7c 1l-,ies of the authors and sho7uld be ; - The finidinzgs, interpretations, and conclusions expressed in this paper are c;itircly thosc of the authors. They do not necessarily represent the viCte of the World Bank, its Executive Directors, or the cou1nties they represent. Produced by the Policy Research Dissemination Center Comments Welcome Governance Matters II: Updated Indicators for 2000-01 Daniel Kaufmann*, Aart Kraay*, and Pablo Zoido-Lobat6n** * The World Bank **Stanford University February 2002 This paper is a product of the ongoing project on worldwide governance research indicators led by Kaufmann and Kraay. We are grateful to Massimo Mastruzzi, Tugrul Gurgur and Bogdan Cosmaciuk for excellent research assistance, and to David Cieslikowski and Steve Knack at the World Bank for inputs, and Marc Levi at Columbia University, Dr. Ted Haner and Steven Weber at BERI, Merli Baroudi at EIU, Meril James at Gallup, John McArthur at C.l.Dat Harvard University, Peter Cornellius at the World Economic Forum, Johann Graf Lambsdorff at Transparency International, and Joel Jurtzman and Carlo di Florio at PWC, for providing access to data and explanations. The authors can be contacted at dkaufmannaworldbank.orp, akraa)-worldbank.org, and pzoido(cD-stanford.edu. The governance research indicators presented in this paper are available electronically at http://www.world bank. orp/wbi/povernance/qovdata200 1 or httP:/Hwww.worldbank. orq/research/p rowth. 1. Introduction The objective of this short paper is to present updated estimates of six dimensions of governance for 175 countries in 2000-01. We draw on 17 separate sources of subjective data on perceptions of govemance constructed by 15 different organizations, and use the unobserved components methodology described in our previous work (Kaufmann, Kraay and Zoido-Lobat6n (1999a,b)) to construct updates of our governance indicators covering the period 1997-98. Table 2 of this paper reports both the 1997-98 and the 2000-01 indicators. The data is also available electronically at http://www.worldbank.org/wbi/governance/govdata2001.htm or http://www.worldbank.org/research/growth. In our previous work, we constructed aggregate governance indicators for six broad concepts of govemance, using an unobserved components methodology that permitted us to also report measures of the precision of governance estimates for each country. One of the main findings of this previous research was the margins of error associated with the composite estimates of governance for each country are typically quite large relative to the units in which governance is measured. This implies that cross-country comparisons of the quality of governance based on this type of data need to be made with considerable caution: many of the small measured differences in govemance perceptions are too small to be statistically - or practically - significant, and only large differences are likely to be statistically meaningful. The updated governance indicators in this paper display the same substantial margins of error as in our previous work - although by drawing on a slightly larger set of underlying sources of governance data, the average precision of the governance indicators is slightly higher. As a result, the same caution about the robustness of cross- country governance rankings stressed in our previous work continues to apply here. Using the updated governance indicators, we can now also for the first time compare changes over time in point estimates of govemance with measures of precision of these estimates. For the majority of countries, changes in governance perceptions over the four-year period we consider are small relative to the margins of error associated with levels of governance. This points to the importance of treating many of the small observed changes over time in governance with the same caution as the many small differences in the level of governance across countries at a given point in time. 1 Nevertheless, we do find that for some countries, changes over time in governance perceptions even over the fairly short four-year period we consider are large relative to associated margins of error, and point to possibly significant governance trends for these countries. The rest of this paper proceeds as follows. In the following section, we describe the updated governance database on which the 2000-01 indicators are based. In Section 3 we briefly review the concepts of governance corresponding to our six aggregate indicators, and the methodology used to construct them. Section 4 describes the levels and changes over time in our new estimates of governance. Section 5 concludes. 2. An Expanded Governance Database In this section we describe the expanded set of measures of governance we have used to construct our six composite governance indicators for 2000/01. In this round, we rely on 194 different measures drawn from 17 different sources of subjective governance data constructed by 15 different organizations, as listed in Table 1. These sources include international organizations, political and business risk rating agencies, think tanks, and non-governmental organizations. Four of these sources are included in the 2000/01 index for the first time, and the remaining are updates of sources included in the 1997/98 indicators.' In this section, we provide an overview of some of the key features of these sources. Appendix 1 presents a detailed description of each of these sources. I Two of the sources we refer to as 'updates of existing sources" are in fact new surveys which included questions similar to those in their previous incarnations. These are the World Bank's World Business Environment Survey which is a follow-up in the 1997 World Development Report Survey, and the Gallup Millenium Survey which we treat as a follow-up of the Gallup 50h Anniversary Survey included in our previous paper. There is also some overlap between sources. A portion of the World Bank's Business and Enterprise Environment Survey (BPS, see htti://lwww.worldbank.org/wbi/aovernance/beeosinteractive.htm) was used in the World Business Environment Survey, and we refer to the former as a distinct source only when we use questions that appeared only in the former. Also, in 1997 we used data from the Global Competitiveness Survey for Africa as a separate source. In 2000, with the exception of the questions on state capture for transition economies, its questionnaire coincided with that of the World Business Environment Survey, and we include it in the latter source. Finally, one source we used in 1997/98 is no longer published (Central European Economic Review), and so we are unable to update this source. 2 As in our previous work, we continue to distinguish sources according to (1) whether they are polls of experts, or surveys of businesspeople or citizens in general, and (2) the extent to which the sample of countries included in the sources is representative of the world as a whole. In the 1997/98 version of the govemance database, our sources consisted of eight polls of experts and seven surveys, and five of our sources were classified as representative. In the 2000/01 version, we have added two new surveys (the World Bank's Business Environment and Enterprise Performance Survey of firms in transition economies (BPS), and Latinobarometro (LBO), a private survey conducted in 17 Latin American countries), and two new polls of experts (the Columbia University State Capacity Study (CUD) covering 109 countries worldwide, and the PriceWaterhouseCoopers Opacity Index (PWC) covering 35 countries). The advantages and disadvantages of polls of experts relative to surveys of market participants are well-known, and are discussed in our previous paper (KKZ 1999b). Briefly, the choice among these two types of governance data involves tradeoffs in terms of cross-country comparability versus first-hand knowledge of local conditions. Polls of experts in general are explicitly designed to provide comparable results across countries, through elaborate benchmarking procedures. However, their reliability depends greatly on the ability of the small group of experts involved to provide objective and accurate assessments of the governance dimensions being measured. In contrast, surveys typically draw on the responses of large numbers of local respondents with direct first-hand knowledge of local conditions. However, to the extent that ostensibly identical survey questions are interpreted differently by respondents with different cultural and/or socioeconomic backgrounds, it can be difficult to make cross- country comparisons using survey responses.2 Sources of governance data also vary with respect to the sample of countries they cover. A number of sources cover a very large sample of developed and developing countries, while others cover very narrowly-focused samples of countries. Also, many of the poorest and/or smallest countries in the world tend not to be covered by many commercially-oriented polls because they are relatively unattractive to foreign investors. 2 For example, it is easy to see how responses to a question on the prevalence of "improper practices" can be affected by country-specific perceptions of what such practices might be. On various approaches to address differences in country-specific perceptions, see Hellman et. al. (2000) and Kaufmann and Wei (1999). 3 Since there is a strong positive association across countries between governance and per capita incomes, this difference between sources makes it difficult to compare indicators from sources which cover sets of countries with very different income levels. Similarly, there may be regional differences in governance which hamper comparisons across sources. For example, it is not clear how to compare a governance rating based only on transition economies with one based on a broad set of countries. As discussed in a previous paper (KKZ 1999a), the methodology we use to construct aggregate governance indicators takes these differences in country coverage into account as the data from individual sources are transformed into common units for aggregation across sources. Although several of the sources we use have somewhat expanded their country coverage, this does not alter our previous classification of sources as representative and non-representative in the earlier version of the indicators. We use the same criteria of representativeness according to geographical location and income levels in order to identify one of our new sources as representative (CUD), and the remainder as non- representative (PWC, LBO, and BPS), for a total of seven representative sources and ten non-representative sources. 3. Methodology In this section, summarizing our work in KKZ (1 999a,b) we first describe how we organize the available govemance data described above into six clusters corresponding to six basic aspects of governance, and describe how we combine these indicators into aggregate governance indicators. Governance Clusters We construct the same six aggregate governance indicators as in our previous work, motivated by the same definition of governance that we used before. In KKZ 1 999b, we defined govemance as the 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 govemment to effectively formulate and implement sound policies, and (3) the respect of citizens and the state for the institutions 4 that govern economic and social interactions among them. We stress at the outset that the classification of indicators into clusters corresponding to this definition of governance is not intended to be definitive. Rather, it reflects our own views of what constitutes a useful and interesting organization of the data that is consistent with prevailing notions of govemance. The first two governance clusters are intended to capture the first part of our definition of governance: the process by which those in authority and selected and replaced. We refer to the first of these as 'Voice and Accountability", and include in it a number of indicators measuring various aspects of the political process, civil liberties and political rights. These indicators measure the extent to which citizens of a country are able to participate in the selection of governments. We also include in this category indicators measuring the independence of the media, which serves an important role in holding monitoring those in authority and holding them accountable for their actions. The second govemance cluster is labeled "Political Stability". In this index we combine several indicators which measure perceptions of the likelihood that the government in power will be destabilized or overthrown by possibly unconstitutional and/or violent means, including terrorism. This index captures the idea that the quality of govemance in a country is compromised by the likelihood of wrenching changes in government, which not only has a direct effect on the continuity of policies, but also at a deeper level undermines the ability of all citizens to peacefully select and replace those in power. The next two clusters summarize various indicators of the ability of the government to formulate and implement sound policies. In "Government Effectiveness" we combine perceptions of the quality of public service provision, the quality of the bureaucracy, the competence of civil servants, the independence of the civil service from political pressures, and the credibility of the government's commitment to policies into a single grouping. The main focus of this index is on "inputs" required for the government to be able to produce and implement good policies and deliver public goods. The second cluster, which we refer to as "Regulatory Quality", is more focused on the policies themselves. It includes measures of the incidence of market-unfriendly policies such as price controls or inadequate bank supervision, as well as perceptions of the burdens imposed by excessive regulation in areas such as foreign trade and business development. 5 The last two clusters summarize in broad terms the respect of citizens and the state for the institutions which govern their interactions. In "Rule of Law" we include several indicators which measure the extent to which agents have confidence in and abide by the rules of society. These include perceptions of the incidence of both violent and non-violent crime, the effectiveness and predictability of the judiciary, and the enforceability of contracts. Together, these indicators measure the success of a society in developing an environment in which fair and predictable rules form the basis for economic and social interactions. The final cluster, which we refer to as "Control of Corruption", measures perceptions of corruption, conventionally defined as the exercise of public power for private gain. Despite this straightforward focus, the particular aspect of corruption measured by the various sources differs somewhat, ranging from the frequency of "additional payments to get things done," to the effects of corruption on the business environment, to measuring "grand corruption" in the political arena or in the tendency of elite forms to engage in "state capture". The presence of corruption is often a manifestation of a lack of respect of both the corrupter (typically a private citizen or firm) and the corrupted (typically a public official) for the rules which govern their interactions, and hence represents a failure of governance according to our definition. Aggregating Governance Indicators Implicit in our organization of the data is the view that, within each cluster, each of these indicators measures a similar underlying basic concept of govemance. Given this view, there are considerable benefits from combining these related indicators into an aggregate govemance indicator for each cluster. First, the aggregate indicators span a much larger set of countries than any individual source, permitting comparisons of govemance across a broader set of countries than would be possible using any single source. Second, aggregate indicators can provide more precise measures of governance than individual indicators. Third, it is possible to construct quantitative measures of the precision of both the aggregate govemance estimates for each country, as well as their components. This allows formal testing of hypotheses regarding cross- country differences in governance. For each of these clusters, we combine the component indicators into an aggregate govemance indicator using the same methodology used to calculate our first 6 set of indicators, as documented in KKZ (1999a). We use an unobserved components model which expresses the observed data in each cluster as a linear function of the unobserved common component of governance, plus a disturbance term capturing perception errors and/or sampling variation in each indicator. In particular, we assume that we can write the observed score of country j on indicator k, y(,k), as a linear function of unobserved governance, go), and a disturbance term, E(,k), as follows: (1) y(j,k) = a(k) + P(k) (g(j) + s(j,k)) where cx(k) and 5(k) are unknown parameters which map unobserved governance go) into the observed data y(,k). As a choice of units, we assume that go) is a random variable with mean zero and variance one. We assume that the error term has zero mean and a variance is the same across countries, but differs across indicators, i.e. E[e(j,k)2 j=2 (k). Given estimates of the parameters of the model, a(k), ,B(k), and c(k), we can compute estimates of governance for each country, as well as measures of the precision of these estimates.3 Formally, the estimate of governance for a country produced by the unobserved components model is the mean of the distribution of unobserved govemance conditional on the Ka) observed data points for that country. This conditional mean is the following weighted average of appropriately-rescaled scores of each of the component indicators: K(j) yjk)- cx(k) (2) E[g(j) I y(j,1),..., y(j,K(j))] = Ew(k) (k) G(k) - where the weights applied to each source k, w(k) = , are inversely 1+ Ya,(k)-2 k=1 proportional to the variance of the error term of that source. We also report the standard 3Estimation of this model requires some non-standard techniques to take into account the likely differences in the unobserved distribution of governance across the different samples of countries covered by representative and non-representative sources. These are documented in detail in KKZ (1999a). 7 deviation of this conditional distribution as an indicator of the confidence we can have in this estimate, which is: (3) SD[g(j) I y(j,1),.. , y(j, K(j))] =1 + E c(k)-2 k=1 This standard deviation is declining in the number of individual indicators in which a particular country appears, and is increasing in the variance of the disturbance term on each of these indicators. The assumptions of the unobserved components model ensure that the distribution of governance in each country is normal, conditional on the data for that country. Therefore, these conditional means and standard deviations for each country have a natural interpretation. For example, a useful interpretation of the reported estimates and standard deviations for each country is to note that there is a 90% probability that the "true" level of governance in a country is in an interval of plus or minus 1.64 times the reported standard deviation centered on the point estimate itself. We refer to such a range as a 90% confidence interval around the estimate of governance for a country.4 Our choice of units for governance ensures that the estimates of governance have a mean of zero, a standard deviation of one, and range from around -2.5 to around 2.5. Moreover, since we adopt the same choice of units for governance in each period, the indicators are not informative about a worldwide average trend in governance. However, they are informative about how countries' relative positions change over time, as discussed further below. The aggregate indicators are oriented such that higher values correspond to better outcomes. 4This is a slight abuse of terminology, as these are not confidence intervals in the usual frequentist sense of a stochastically varying interval centered around a fixed unknown parameter. Rather, we treat governance as a random variable, and the 90% confidence interval is simply the 5th and 95th percentiles of the conditional distribution of governance given the observed data. 8 4. Results In this section we first describe the updated governance indicators for 2000/01, and then discuss changes in these governance indicators relative to the 1997/98 version of the indicators. Cross-Country Differences in Governance in 2000/01 Table 2 reports the point estimates of governance, the estimated standard errors, and the number of sources by country, for each of six governance indicators, for 2000/01. For reference, the corresponding information for the 1997/98 indicators are reported as well in the same table. Table 3 summarizes the number of countries and sources included in each indicator. For our indicators, the addition of new sources of governance data results in only a small increase in the number of countries covered. As emphasized in our previous work, we find that the six dimensions of governance are not very precisely measured, in the sense that the measured standard deviations are large relative to the units in which governance is measured. We illustrate this point in Figure 1. In each panel, we order countries in ascending order according to their point estimates of governance on the horizontal axis, and on the vertical axis we plot the estimate of governance and the associated 90% confidence interval described above. The size of these confidence intervals varies across countries, as different countries appear in different numbers of sources with different variances. The resulting confidence intervals are large relative to the units in which governance is measured. To emphasize this key point, the horizontal lines in Figure 1 delineate the quartiles of the distribution of governance estimates. Relatively few countries have 90% confidence intervals that lie entirely within a given quartile. From these figures, it is also clear that many of the small differences in estimates of governance across countries are not likely to be statistically significant, mirroring the reality that it is in fact difficult to distinguish small differences among countries using this type of data on governance. For many applications, it is therefore more useful to focus on the range of possible governance for each country as summarized in the 90% confidence intervals shown in Figure 1. For two countries at opposite ends of the scale 9 of governance, whose 90% confidence intervals do not overlap, it is clear that there are in fact significant differences in governance between these two countries. For pairs of countries that are closer together and whose 90% confidence intervals overlap, one should be much more circumspect about the significance of estimated differences in governance between two such countries. An important consequence of this imprecision is that it is difficult to use only the point estimates of governance to sort countries into even relatively broad governance categories. Rather, it seems more appropriate to identify groups of countries with extreme governance outcomes by explicitly taking into account not only the point estimates of governance, but also the substantial margins of error associated with individual governance scores. For example, it seems more plausible to identify countries with clear governance failures as those for which the associated 90% confidence interval lies entirely within the bottom third of the point estimates of governance, and countries with a clear absence of governance problems as those for which the associated 90% confidence interval lies entirely within the top third of the point estimates of governance. Table 4 reports the number of countries falling into each of these groups, and the fraction of all countries that they represent, for each of the six measures of govemance. This table shows that slightly under half of the countries with point estimates of govemance in the bottom third of countries also have 90% confidence intervals that fall entirely within this bottom tercile, for each of the six governance indicators. In addition, somewhat more than half of the countries in the top tercile also have 90% confidence intervals entirely within the top tercile. The fraction of countries significantly in the top tercile is slightly larger than that in the bottom quintile because these countries tend to appear in more of our sources of governance data, and so typically have smaller standard errors5. Despite the imprecision of these aggregate indicators, we argue that they are still useful, for several reasons. First, since each of these aggregate indicators spans a much larger set of countries than any individual indicator, it is possible to make 5Another way of segmenting the data also suggests that identifying groups of countries with extreme governance outcomes can be statistically meaningful: over 85% of the countries in the bottom quartile have 90% confidence intervals entirely within the bottom 40 percent of the distribution of governance estimates; for the top quartile the share of countries (in the top 40 percentile) exceed 95 percent. 10 comparisons - however imprecise -- across a much larger set of countries than would be possible with any single indicator.6 Second, although imprecise, each aggregate indicator provides a more precise signal of its corresponding broader governance concept than do any of its component indicators, and moreover provides a convenient and consistent summary of the available evidence. Third, the measures of precision for each country are useful because they enable formal statistical tests of cross-country differences in governance instead of arbitrary comparisons. Fourth, and finally, it is worth noting explicitly that the measurement error we have documented in these governance indicators need not make them unsuitable for cross-country econometric analysis. Rather, one possible advantage of these indicators is that one can use information in the estimates of the precision of each aggregate to quantify the effect of measurement error in regression analyses that use governance indicators as right-hand side variables.7 Finally, Table 5 reports our point estimates of the parameters of the unobserved components model, for each of the six governance indicators in 1997/98 and in 2000/01. The most interesting of these are the estimates of the standard deviation of the error term of the individual sources, since as discussed above the weights used to combine individual sources into aggregate indicators are inversely proportional to the squares of these standard deviations. Figure 2 presents a visual summary of how the weights applied to individual sources vary across sources, indicators, and time periods. For now, we simply note that there are some non-trivial differences between the 1997/98 and the 2000/01 exercise in how individual sources are weighted, reflecting changes over time in the estimated precision of such individual sources. In the next subsection, we discuss more systematically the contribution of these changes in weights to changes over time in the aggregate indicators. 6 See for example Knack and Azfar (2000) who document the importance of sample size for conclusions about the determinants of corruption. 7See KKZ1999b for regression analysis of the effects of these governance indicators on socio- economic variables, and a discussion of the consequences of measurement error. 11 Changes in Governance 1997198 to 2000/01 We now tum to changes over time within countries in our estimates of the six dimensions of governance. To give a visual perspective on these changes, Figure 1 also indicates the score of each country on the corresponding 1997/98 indicator, as a heavy dot. Two observations on these figures are noteworthy. First, the correlations across countries between the 1997/98 and 2000/01 indicators are quite high, as evidenced by the strong upward trend in the heavy dots moving from right to left across countries sorted by their 2000/01 scores. The last column of Table 3 reports these correlations, which are all greater than 0.9. Second, despite these strong correlations reflecting a high degree of persistence in perceptions of countries' relative govemance quality, there are substantial changes for some countries. By substantial, we mean that the change in a country's estimate of governance between 1997/98 and 2000/01 is large relative to its margin of error for 2000/01. While the exact definition of what constitutes "large" is of course arbitrary, one standard is to look at countries for which the 1997/98 estimate of governance falls outside the 90% confidence interval for the 2000101 score. In Figure 1, countries with significant changes in estimates of governance are depicted by those dots (1997/98 values) located outside the 90% confidence range for the 2000/01 estimates. Consulting the 1997/98 and 2000/01 scores and corresponding standard errors in Table 2, we find a number of substantial changes for each governance cluster. For instance, Croatia, Bosnia, Nigeria, Indonesia, Peru, Slovakia and Ghana are some of the countries exhibiting significant improvements in Voice and Accountability during the period, while by contrast the data suggests that Ivory Coast, Venezuela, Congo, Belarus and Pakistan have deteriorated in this particular dimension. Improved regulatory regimes are apparent in countries like Singapore, Azerbaijan, Albania, Thailand, Botswana and Estonia, in contrast with countries such as Zimbabwe, Indonesia, Belarus, Congo Democratic Republic (ex- Zaire), Argentina, Russia and Venezuela. On Government Effectiveness, countries like Estonia, Tunisia, Botswana, Bulgaria, Mauritius, Namibia, Dominican Republic, and Ireland are shown to be improving in their government effectiveness dimension, contrasting with countries like Sierra Leone, Peru, Mali, Ivory Coast and Ecuador. 12 For Political Stability, Brazil, Haiti, Libya, Algeria and Nicaragua illustrate improvements, in contrast with countries like Ivory Coast, Tanzania and Zimbabwe. On Rule of Law, Belgium, Bosnia, Mozambique, Estonia, Uruguay, Latvia and Slovakia exemplify improvements, contrasting the evidence on deteriorating rule of law aggregate indicator in countries like Zimbabwe, Moldova, Malaysia, and Philippines. Finally, on control of corruption, improvements are exemplified by Namibia, Tunisia, Croatia, Dominican Republic, Chile and Costa Rica, in contrast with countries like Zimbabwe, Malaysia, Kenya, Russia, and the Ivory Coast.8 From these examples, it is apparent that changes in the various govemance components do tend to move together on average, i.e. there is a positive correlation across changes for a country across components. A country like Zimbabwe, with across-the-board deterioration in governance, illustrates such a correlation across indicators. Yet for the worldwide country sample such correlation, while positive, is far from perfect, because for many countries the various governance indicators do not move in tandem: a country can exhibit improvement in one dimension and deterioration in another. For instance, Indonesia exhibits major improvement in Voice while Regulatory Burden, Political Stability, and Control of Corruption have deteriorated, while in Nigeria Voice has improved but Political Stability has worsened.9 What accounts for the observed changes in estimates of governance for individual countries between 1997/98 and 2000/01 ? Mechanically, changes over time in the governance indicators are due to a combination of three factors: (i) changes in governance perceptions as reported by the individual underlying sources; (ii) changes in the weights the aggregation procedure assigns to the individual underlying sources; and 8 In reviewing which countries have undergone a significant change over the period, we note that some large positive changes take place from an initially low level, and vice-versa regarding some large negative changes. Any in-depth trend analysis for a country ought to also take into account its initial level estimates. 9It is worth noting that while overall the data appear to make sense when confronted by what we know about the level and trends in governance in most settings, at the same time it ought to be noted that not al/ the changes that the data exhibits would necessarily concord with one's priors. In part this is a reminder that oddities in the data can take place even with aggregate indicators, which in turn is a reminder of the potential piffalls of relying excessively on any individual indicator. It is also important to note that the aggregate indicators are obtained on the basis of data collected in 2000-01, with the cutoff having been in August 2001. Thus, while the data is relatively recent, these indicators do not capture the most recent developments in govemance. 13 (iii) changes in the set of underlying sources used to construct the aggregate governance indicators. We now provide a systematic accounting of the relative importance of each of these three factors. To do so, note that we can write the change between period t and period t-1 in country j's score on each of our governance indicators as: (4) jjt9j- L,t- J1j_ ]+ [ajt j,t-i t-,-9,t ) whereg 1t denotes the estimate of governance in country j in period t, and gjt denotes the estimate of governance in country j in period t based using data only from a "balanced" set of sources and countries for which data is available in both periods. The first term in square brackets on the right-hand side of Equation (4) therefore represents the change over time in estimated governance using only data from the same countries and same sources in both periods. The second term represents the contribution of changes in underlying sources and country coverage to the change in the governance indicator, i.e. item (iii) above, which consists of the change in the difference between country j's score using all available data and using only data available in both periods. The first term in Equation (4) above can be further decomposed to isolate the effect of changes over time in the weights assigned by the aggregation procedure, as follows: '-B ' B [6BO, _ B I J '+ [1t_B0j (5) 9j,t 9J,t-= bj,t -j,t_1+9 _ gBS where gj,tO denotes an estimate of governance in period t using only the "balanced" data as before, but also using the "old" weights used in the base period to aggregate sources. The first term on the right-hand side of Equation (5) therefore reflects the change in governance estimates attributable solely to changes in the country's score on each of the underlying individual indicators, i.e. item (i) above, while the second term reflects the change due to the re-weighting of the individual sources, i.e. item (ii) above. 14 Clearly the contribution of each of these three factors to changes in estimates of governance over time will be different for each country. However, we can summarize the overall importance of each of these three considerations with the help of a simple variance decomposition. In the first column of Table 6, we report the standard deviation across countries of changes in our governance indicators, for each of the six dimensions of govemance we consider. In the second column, we report the share of the variance of changes in governance that is attributable to the first term in Equation (4) above, i.e. the fraction of the variance of changes over time due to changes in the "balanced" indicator which uses only data from the same sources and countries in both periods.10 In the third column, we report the share of the variance of changes in the "balanced" governance indicators that is due to the first term in Equation (5) above, i.e. the fraction of the variance of changes over time in the "balanced" indicator due only to changes over time in the assessments of individual sources. The fourth column of Table 6 reports the product of columns 2 and 3, which can be interpreted as the fraction of the variance of changes in the overall governance indicator that is due only to changes over time in the assessments of individual sources. For most of our indicators, we find that over three-quarters of the variance in changes in scores over time can be attributed to changes in scores on individual sources (with the exception of Rule of Law (0.57), and Control of Corruption (0.43). For these last two indicators, changes in weights and the addition of new sources accounts for a substantial fraction of the variation in changes in governance over time. 10 In this table, we follow the convention of assigning half of the covariance between the two terms to the share of the variance attributable to the first term (see Klenow and Rodriguez (1999) for a justification of this convention). A disadvantage of this rule of thumb is that it is possible for the fraction of the variance attributable to the first term to exceed one. Note also that the covariance terms in each case are typically an order of magnitude smaller than the variance terms, so the precise convention for assigning the covariance matters little for the overall conclusions. 15 Conclusions In this short note, we have constructed aggregate governance indicators for six dimensions of governance, spanning 175 countries in 2000/01. These new indicators update our previous governance indicators for 1997/98. By drawing on a larger set of sources, we have been able to slightly expand country coverage and slightly improve the precision of the aggregate indicators. Nevertheless, we find that - as emphasized in our previous work - margins of error associated with estimates of governance remain substantial. This points to the importance of caution in making comparisons across countries and over time along these various dimensions of governance. Nevertheless, we do find that the indicators are sufficiently informative that we can confidently identify broad groupings of countries with extremes of governance. We also find that for a number of countries, changes in governance perceptions even over the relatively short four-year period we consider are large relative to measures of imprecision, pointing to possibly significant governance trends in these countries. This in turn opens the possibility of future cross-country empirical work to understand the causes and consequences of these changes in governance overtime. 16 References Hellman, Joel, Geraint Jones, Daniel Kaufmann, and Mark Schankerman (2000). "Measuring Govemance, Corruption, and State Capture: How Firms and Bureaucrats Shape the Business Environment in Transition Economies". World Bank Policy Research Working Paper No. 2312. http://www.worldbank.org/wbi/lovernance/wp-statecapture.htm Kaufmann, Daniel, Aart Kraay and Pablo Zoido-Lobat6n (1999). "Aggregating Govemance Indicators". World Bank Policy Research Department Working Paper No. 2195. http://www.worldbank.orq/wbi/governance/pubs/acgindicators.htm Kaufmann, Daniel, Aart Kraay and Pablo Zoido-Lobat6n (1999). "Governance Matters". World Bank Policy Research Department Working Paper No. 2196. http://www.worldbank.org/wbi/governance/pubs/covmatters.htm Kaufmann, Daniel and Shang-jin Wei (1999). "Does 'Grease Money' Speed Up the Wheels of Commerce?". NBER Working Paper No. 7093. http://www.worldbank.orq/wbi/Qovemance/pubs/Qreasemoney.htm Klenow, Peter J. and Andres Rodriguez-Clare. "The Neoclassical Revival in Growth Economics; Has It Gone Too far?" in Ben S. Bernanke and Julio Rotemberg eds. NBER Macroeconomics Annual 1997. Knack, Stephen and Omar Azfar (2000). "Are Larger Countries Really More Corrupt?" World Bank Policy Research Department Working Paper No. 2470. Levi, Marc (2000) ."Measuring State Capacity". Mimeo, Columbia University. Note: Descriptions of the publications in which our various sources of govemance data appear are provided in the corresponding Appendix tables. 17 Table 1: Sources of Governance Data, 2000101 Country Repre- New in Source Publication Code Tvpe Coverage sentative 2000/01 Business Environment Risk Intelligence Business Risk Service BRI Poll 50 Columbia University State Capacity Project CUD Poll 109 x x Economist Intelligence Unit Country Risk Service EIU Poll 115 x European Bank for Reconstruction and Redevelopment Transition Report EBR Poll 26 Freedom House Nations in Transition FHT Poll 27 Freedom House Freedom in the World FRH Poll 192 x Gallup International Gallup Millennium Survey GMS Survey 60 Heritage Foundation/Wallstreet Journal Economic Freedom Index HWJ Poll 161 x Institute for Management and Development World Competitiveness Yearbook WCY Survey 49 Latinobarometro Latinobarometro Surveys LBO Survey 17 x Political Economic Risk Consultancy Asia Intelligence PRC Survey 14 Political Risk Services International Country Risk Guide PRS Poll 140 x PriceWaterhouseCoopers Opacity Index PWC Survey 35 x Standard and Poor's DRI McGraw-Hill Country Risk Review DRI Poll 111 x World Bank Business Enterprise Environment Survey BPS Survey 18 x World Bank World Business Environment Survey WBS Survey 81 x World Economic Forum Global Competitiveness Report GCS Survey 75 Table 2: Estimates of Governance Voice and Accountability Political Stability Govemnment Effectiveness 2000/01 1997198 2000101 1997/98 2D00001 1997/98 Est S.E. N Est. S.E. N Eat S.E. N Eat. S.E. N Eat. S.E. N Est S.E. N AFG AFGHANISTAN -1409 1 -1.62 0.36 1 -.3 49 1. . -4. *- 1 ALB ALBANIA .9 017 8-S.: -0.13 0.21 4 -0 0 037 4 -1.00 0.25 4 -089 <025 5 -0.65 0.29 4 DZA ALGERIA -129".(4 2 1.31 0.25 3 -1 27 028 4 -2.42 0.30 3 .-081 0 24 4 -1.09 0.32 3 AGO ANGOLA .1,20 0.24 4 -1.~00 0.25 3 -1 98 0.28 4 -1.78 0.28 4 AM 0.24 4 -1,39 0.24 5 ARG ARGENTINA 057 023 ,7.0.49 0.25 4 0 55 022 9 0.51 0.26 5 0.18 0.18 '10 0.26 0.25 6 ARM ARMENIA .022 0 117S. -0.18 0.21 4 -0 84 0.41 4 -0.45 0.25 4 .1 03 0.277 5 -0.65 0.29 4 AUS AUSTRALIA I 10 0124 51:1 1.63 0.25 4 120 0 23 7 1.18 0.26 5 1 58 0 19 7 .1.46 0.25 6 AUT AUSTRIA 1 34 0 24 5. 1.45 0.25 5 127 0.23 7 1.38 0.25 6 1 51 0 20 7 1.22 0.23 7 AZE AZERBAIJAN -0170 0.16 6 -0.81 0.19 5 -0 70 0 27 5 -0.36 0.23 5 0 96 0 21 6 -0.83 0.24 5 BHS BAHAMAS 115 0.28 2 1.13 0.29 2 0.68 085 1 0.37 0.41 1 10D4 053 1 0.47 0.77 1 BHR BAHRAIN -0 96 0.24 4 -1.04 0.25 3 -0 04 0 21 4 -0.08 0.30 3 0 62 0 24 4' 0.24 0.32 3 BGD BANGLADESH -0 20 0.224 5 . -0.01 0.25 3 .067 021 5 -0.40 0.30 3 -054 0 23 5 -0.56 0.32 3 BRB BARBADOS 1.27 0.39 1 1.49 0.36 1 -. BLR BELARUS -1 04 0.17 5 -0.78 0.21 4 004 037 5 -0.37 0125 4 .099 0 25 6 -0.66 0.29 4 BEL BELGIUM 1.24 0.24 5,~ 1.41 0.25 4 057 0.23 7 0.82 0.26 5 1 29 0.20 1 0.88 0.25 6 BLZ BEUIZE 1 01 0.39 :21,23 0.36 1 0320 1 055 oS 070 BEN BENIN 047 0 36 2. 0.89 0.35 2 .-72 0 63 1 -0.94 0.66 1 012 0.52 ½ -0.07 0.56 1 BTN BHUTAN 1 27 0 39 f;. -1.25 0.36 1 BOL BOUIVIA 0 21 0 24 .- 0.39 0.25 4 -0861 0 28 6 -0.14 0.28 4 .0 47 0 22 6 .-0.22 0.26 5 BIH BOSNIA-HERZEGOVINA .02* 0.9 4 -1.00 0.23 2 .-0`01 0.84 2 -1.16 0.39 1 49'0.39 3 -1.11 0.41 1 BWA BOTSWANA .0 024 .4) 0.78 0.25 3 071 02? 5 07 .8 4 00- 0 23 5 0.22 0.26 4 BRA BRAZIL 0 53 0 24' 0.58 0.25 5 047 0 22 '* -0.32 0.25 6 027 OA1 9 -0.22 0.23 7 BRN BRUNEI -0 93 0.28 '--0~92 0.29 2 '386 085 01. 132 1A 1 0.68 0 53 1 0.01 0.77 1 BGR BULGARIA 0O.59 0.116 1 0.47 0.19 5 GS?- SS 0.43 0.23 5 .0 28 0.19 B -0.81 0.24 5 BFA BURKINA FASO -0 26 02Z7 $.3-0.21 0.29 2 .0 54 051 2 -0.52 0.36 2 -0.02 0 40 2 -0.06 0.32 3 BDI BURUNDI -¶ 35 0 36 .22.] -1.29 0.36 1 -i154 0.54 2 -1 14 0.37 2 KHM CAMBODIA -0.17 0 392 -0.91 0.36 1 -0 13 0 90 1 -0.34 070 1 CMR CAMEROON -0.82 024 5 -0.70 0.25 4 -0 13 0 29 5 -0.72 0.27 5 .40.# 025 5 -0.64 0124 5 CAN CANADA -0.3 *13~ 7 1.39 0.25 5 124 0 22 9 1.03 0.25 6 1711 019S 1.72 0.23 7 CRyV CAPEVERDE 002 0 39 1 0.99 0.36 1 , ".. . CAF CENTRAL AFRICAN REPUBLIC -09 09 1 -.5 03 ~ -~. " '-0.75 0.51 1 TCD CHAD -0898 0 39 1 -0.85 0.35 2 ,-0.74 0.66 1 . . -0.71 0.56 1 CHL CHILE 0.63 0 23 7 0.62 0.25 4 04'O$ 0.22 9 0.45 0.26 5 1 13 018S 10 1.17 0.25 6 CHN CHINA -111 0 24 6 -1.29 0.25 5 -03 0 22 8 0.48 0126 5 .0.14 018 9 0.02 0.25 6 COL COLOMBIA -0 41, 0.23, 7 -0.15 0.25 5 '4.36' 0.22 9 -1.29 0.25 6 -0 38 0.18 10a -0.06 0.23 7 COM COMOROS -0.35 039 1, 0.06 0.36 1 ,,.' . COG CONGO -1.38 027 3 -0.77 0.29 3 -13 0' 51 2 -1.83 0.37 2 .158 0'40 Z , -.58 0.39 3 CR1 COSTA RICA 1 37 024 5 1.35 0.25 4 1.08 0.28 6 0.91 0.27 5 0 74 021 6 0.55 0.29 4 CIV IVORY COAST 119g 0.25 4 -0.57 0.25 4 -0 95 029 4 -0.14 0.27 5 .0681 0 25 4 -0.18 0.24 5 HRV CROATIA 0 48 GIG St -0.23 0.19 4 0,18 0.29 5 0.41 0.24 4 010 0 22 6 0.15 0.26 4 CUB CUBA .1 40 0 25 3 -1.46 0.25 3 00? 0 28 4 0.19 0.30 3 -0.2 0 24 4 -0.50 0.32 3 CYP CYPRUS 1 28 0 25 3 1.11 0.25 3 0.48 0.29 3 0.38 0.30 3 0 91 0.26 3 1.04 0.32 3 OZE CZECH REPUBLIC I 04 018f 7 1.20 0.19 6 0.74 0 23 8 0.81 0.22 7 0.583 018 10 0.59 0.21 8 DNK( DENMARK 1 60 0 23 6 1.63 0.25 4 1.34 0 23 8 1.29 0.28 5 1802 0 19,8 1.72 0.25 6 DJI DJ U-4.4 0 39 I -0.60 0,36 1 . - 0DOM DOMINICAN REPUBLIC 31)42 3.24 5 -0.08 0.25 3 0.46 0.28 6 0.12 0.34 2 :-02*i '022 6 -0.83 0.37 2 ECU ECUADOR -~0.114 0 24 5' 0.27 0.25 4 -0 80 0 25 7 -0.47 0.26 5 -0 94 0 20 8 -0.56 0.27 5 EGY EGYPT :-0.65 0124 5 -0.67 0.25 3 021 0 24 7 -0.07 0.25 6 027 01e 8 -0.14 0.22 6 SLV EL SALVADOR 0.21 0.24 5 -0.10 0.25 3 062 0.30 5 -0.02 0.32 3 0.025 0.23 5 -0.26 0.37 2 GNQ EQUATORIAL GUINEA -1 30 0.39 1 -1.39 0.36 1 '. - ERIJ ERITREA -1 04 036 2 -0.59 0.38 1 -0.38 0863 1 -. . . EST ESTONIA 0.94 0.18 8: 0.86 0.19 5 01 2 7 2 8 1 .6 0.24 5 ETH ETHIOPIA -0 85 0.2 4 -0.50 0.29 2 '4055 049 3 0.14 0.36 2 -101 0 37 3 -0.15 0.32 3 FJI FUJI 0 05 0 36 2- 0.01 0.35 2 .039 063 1 0.01 0.66 1 0 38 052 1 0.63 0.56 1 FIN FINLAND 169 0.2 8. 1.63 0.25 4 1.61 0.23 6 1.51 0.26 5 1867 019 6 1.63 0.25 6 FRA FRANCE 1.11 0.23 1 1.15 0.25 5 1 04 0.22 9 0.65 0.25 6 1 24 0.19 8 1.28 0.23 7 GAB GABON -040 0.24 4 -0.31 0.25 3 -0.44 0 28 4~ -0.58 0.34 2 -Q.41 0 24 4 -1.13 0.37 2 GMB GAMBIA -0173 028 2 .-0.97 0.29 2 049 0.85 I 0.56 0.41 1 0.41 0.53 1 0.16 0.77 1 GED GEORGIA -0 07 0.19 5 -0.25 0.23 3 . -1 00 041 4 -0.78 0.31 3 .0.72 0 27 5 -0.51 0.30 3 DEU GERMANY 4t42 0.28;''? 1.46 0.25 5 121 0 22 9 1.32 0.25 6 Dig'01 8 1.41 0.23 7 GHiA GHANA 0.2 I2 -0.43 0.25 4 4I; 0.29.' 5" -0.10 0.27 5 .0.28 S: -0.29 0.22 6 Note: 'Est' refers to the point estimate of governance. "S.E.' refers to the standard error. "N" refers to the number of sources in which the country appears. Governance indicators are oriented so that hig her values correspond to better outcomes, on a scale from -2.5 to 2.5. These ratings are based on subjective assessments from a variety of sources, are subject to substantial margins of error as indicated, and in no way reflect the official view of the World Bank, its Executive Directors, or the countries they represent. 19 Table 2: Estimates of Governance, Cont'd Voice and Accountability Political Stability Government Effectiveness 2000/01 1997198 2000101 1997198 2000/01 1997198 Eat S.E. N Eat. S.E. N Est. S.E. N Est. S.E. N Eat. S.E. N Eat S.E. N GRC GREECE 1.12 0.24 4 1.05 0.25 4 Ofl .4 0.21 0.28 5 0.53 0.20:7 0 58 0.25 6 GTM GUATEMALA -0-33 024 5 -0.56 0.25 3 47 .3 -0.75 0.32 3 45-0.27 . -0.23 0.37 2 GIN GUINEA 498 0.2 2 -0.87 0.29 3 .-099 0.651 1 -1.03 0.37 2 0.41 0.03~ I -0.03 0.51 2 GNB GUINEA-BISSAJ -0.7 0.Z7 3 -0.45 0.29 3 -1tIt, 0.01.4 2 -1.20 0.37 2 -41.40 0.40 Z2 -0.33 0.51 2 GUY GUYANA 10.94- C27 3 1.01 0.29 2 4.00 0*14 .2 -0.20 0.41 1 0.02 0 40 2 0.01 0.77 1 HTI HAITI .08 02 4 -0.71 0.29 2 .0.30 0.49 -4.. -1.71 0.41 1 .43.2 0.37 3 ,-1.23 0.77 1 HND HONDURAS -0.0 -024 5 -0 06 0.25 3 0O20. 0.00> .5 0.33 0.32 3 .01 023 5 -0.41 0.37 2 HKG HONG KONG 431 o.m1 5 aoi1 0.25 6 11.13 0.2 6 0.92 0.27 5 1.120 4-.212 1.25 0.25 6, HUN HUNGARY 44-: 0.16 6 1.22 0.19 6 0 75 0 22 9 1.25 0.22 7 060o 0.17 1 1 0.61 0.21 8 ISL ICELAND 1A3 0.2 4 1A47 0.29 3 1.0 ~040 4 1,25 0.37 2 133 0.33 4 1.50 0.41 3 IND INDIA 0.86 0.2 6 0.36 0.25 6 _-0.05.022. -0.04 0.25 6 -0.17 01.18 9 -0.26 0.23 7 ION INDONESIA 4k40 0.24 6 -1.13 0-25 5 -1.54 022 &, -1 29 0~26 5 .0.50 018 9 -0.53 0.25 6 IRN IRAN 410- 0.24 4 -0.56 0.25 3 0.02 0.25 S 0.13 0.28 4 -021 0.22 5 -0.34 0.29 4 IRa IRAQ -1.93 0.24 4 -1.75 0.25 3 .4.0 O.28 4- -2.24 0.30 3 .1.41 0 24 4 -1.88 0.32 3 IRL IRELAND 1.57 0.2 8 1.53 0.25 5 1_24.- 023 .-g 1.43 0.25 6. 1.79 0,1 f~ ,$ 1 36 0.23 7 ISR ISRAEL 0.98 0 24 5 1.06 0.25 4 -0 54 0.24 .0 -0.46 0.28 4 0.07 0 20 7. 0.69 0.27 5 ITA ITALY 1.10 0.23 7 1.28 0.25 5 0.82 0.22 z.9 1.16 0.25 6 0.41 018 10 0.77 0.23 7 JAM JAMAICA 0178 025 3 0.75 0.25 4 01Oct3M 3. -0.34 0.32 3 -05 0 26 S3. -0 48 0.33 3 JPN JAPAN 1.03 G23S 6 1.14 0.28 4 1.20~ .0 8 1.15 0.29 4 0.93- 019 29R 0.84 0.31 5 JOR JORDAN 0.10 0.24 4 0.15 0.25 4 0.15 0.27 -0 -0.06 0.27 5 0.42 0 22 -5 0.63 0.26 5 KAZ KAZAKHSTAN -0.8 O16 7 -0.69 0.19 5 0.29-0.25 7 0.22 0.22 6 - 0.61 0.1* 09-0.82 0.23 6 KEN KENYA -0.8` 024 5 -0.70 0.25 4 -0.80 0.27' 0 -1.10 0~27 5 4017 0.22 :0' -0.90 0.22 6 PRK KOREA, NORTH -1.82 0,28 2 -1.79 0.29 2 0.79' 01~ 2 0.37 0.41 1 -4,05 0.40 <2, -0.30 0 77 1 KOR KOREA, SOUTH 0.9 0.23 6 0.91 025 6 01 ill* .6 02 04 .9 9 0.41 0.23 7 KWT KUWAIT 0.08 0.25 3 0.00 0.25 3 0.84 0.29- 3 0.68 0.30 3 O.43'~ 0.20' 5 3 -0.06 0.32 3 KGZ KYRGYZ REPUBLIC *8.57 0.19 3 -0.35 0.23 3 -0.32 0b43 2 0.32 0.31 3 -.0,1 0.31 S~ -0.58 0.30 3 LAO LAOS *1.06 0.36 2 -1.05 0.36 1 0.0 .8 1. 0-052 t LVA LATVIA 0.81 0.18 5 0.75 0.19 5 0*60 0.28 . 0 0.46 0.23 5 0.22 021."0' 00 0.24 5 LBN LEBANON -.2 0.24 4 -0.40 0.25 3 'AU0 0.20 4 -0.25 0.30 3 -00'-02 - 0.17 0.32 3 LSO LESOTHO 40.16 0.3 1 -0.15 0~36 1 . :.. -0.82 0.61 1 .. . , -0.46 0.33 2 LBR LIBERIA -1.04 - 027 3 -0.89 0.29 2 -065 0t-+..z -0.95 0.41 1 -0.94 0.4' 2 -0.92 0.77 1 LBlY LIBYA -1.38 0.28 3 -1.35 0.25 3 40 0.201-31 -1.17 0 30 3 .1 0.1 $ -1.32 0.32 3 LTU LITHUANIA 1.00 0.18 7 0.88 0.19 5 029 0.26 - 7, 0.35 0.23 5 0.26 - i7i*. 9 -1 0.13 0.24 5 LUX LUXEMBOURG 1,41 02? 4 1.49 0.29 3 1l.48 043 3~' 1.40 0.37 2 1.3 04 5 1.67 0.41 3 MKD MACEDONIA, FOZRME R 'uGOS 0.0 048 4 0.09 0.21 4 44_,4 O0* 2 -0.40 0.31 3 40A3- 0.8> 5 -0.58 0.27 3 MDG MADAGASCAR 02 027 4 0.31 0.29 3 43 <.9 09 03 0& 03 -) -.9 03 MWI MALAWI 414 0.24 8 0.06 0.25 4 0.3 % 4 00G4 0.29 4 4077f-t -0.62 0.24 5 MYS MALAYSIA -0.132 023 7 -0.09 0.25 6 -0.1 02 9.0.55 0.25 6 0.03" .; . 9 0.71 0.23 7 MDV MALDIVES 481? 030 1 -0.91 0.36 1 , MLI MALI 0.32 -0.27 3 0.42 0.29 3 -013 0151 .12-0.29 0.37 2 .1 44 0.40 2 -0.05 0.51 2 MLT MALTA 1.4.3 0.28 2 1.41 0.29 2 5.0 063 1 .32 0.41 1 0.73 053. 1 0.63 0.77 1 MRT MAURITANIA 4860OM62 -0.97 036 1 '437½0,43 . 0 65 0.52 1 MUS MAURITIUS 1.27 0OM 3 1.01 0.29 3 J142., _5 3 1.14 0.39 3 0.76 0 28 3 0.17 0.27 3 MEX MEXICO 0.12 0.2 7 -0.11 0.25 5 0.6M 02 9 -0.35 0.25 6i 0.23 0.1-. 10 0.18 0.23 7 MDA MOLDOVA 0.12 0.86 6 0.11 0.19 5 -0 29 orV 5_ 0.20 0.23 5 4.10- 0.21 8 -0.46 0.24 5 MNG MONGOLIA 0.73 0.27 3 0.63 0.21 3 02i 0r.01 -2 0.37 0.41 1 0.80. 0.40 2 1-0.02 0 39 2 MAN MOROCCO -0.23 a,2 3 -0.24 0.25 4 0.16 - 029- 3~ 0.09 0.27 5 040' 7 0 28 3 0.27 0.22 6 MOZ MDZAMBIQUE 422 0.27 3 -0.17 0 29 3 026 06 2 -0.53 0.34 3 4~49>0 * -0.33 0.29 4 MMR MYANMAR -in9 0.24 4 -1.75 0.25 3 -1 20 '028 <4` -0.97 0.30 3 -4.5 0.- 6 ?4` -1.46 0 32 3 NAM1V NAMIBIA 0.32 0.24 5 0.47 0.25 3 :462 QS31 4~ 0.71 0.31 3 0.00 0.20.-_A4 0.04 0.26 4 NPL NEPAL 406 0.36 2 0.05 0.36 1 .0.26 0.3 I . . C .2 1 N.D NETHERLANDS tit OM2 6 1.64 0.25 4 1,45 0.2' 1.48 0 26 5 1.84 ~0.14 -# 2.03 0.25 6 NZL NEW ZEALAND 1859 0.2 5 1.47 0.25 4 .1X~.2. 0.%J4 0 1.42 0.28 4 127 0 21 6 1.57 0.27 5 NIC NICARAGUA -OO-. .024. 5 0.07 0.25 3 0.31 0 30 5 -0.32 0.32 3 _0 73 023 5 -0.55 0.37 2 NER NIGER 9.11 , 0.27 3 -0.74 0.29 2 -0.81 0.51 2 -0.76 0.41 1 --I18 0 40 2 -1.39 0.77 1 NGA NIGERIA -0.44 0.23 6 -1.23 0.25 4 4-1.30 0.28 7 -1.05 0.27 5 .100D 0 21 7 -1.32 0.22 6 NOR NORWAY 1.8 0.24 5 1.67 0.25 4 . 1.32 023 7. 1-41 0.26 5 1.35 0 20 7 1.67 0.25 6 0O1N OMAN 450.0.5 3 -0.57 0.25 3 1.00 0.28 04..91 0.30 3 085 0 24 4- 0.90 0.32 3 PAK PAKISTAN -1.43' 0.2A 5 -0.44 0.25 3 -0 39 028 ~-6>-0.65 0,26 5 .041 0 22 4,,-0.74 0.26 5 PAN PANAMA 0.77 0.36$ 5 0.66 0.25 3 017 0 26 #: 0.15 0,30 3 -014 0 21 1>-0.28 0.32 3 PNG PAPUA NEW GLONE A .0.08 024 4 0.22 0.25 3 .4.48 0.J8 4t, -0.40 0,34 2 0.07, 0.24.', -0.69 0.37 2 Note: 'Est' refers to the point estimate of governance. "S.E." refers to the standard error. "N" refers to the number of sources in which the country appears. Governance indicators are oriented so that higher values correspond to better outcomes, on a scale from -2.5 to 2.5. These ratings are based on subjective assessments from a variety of sources, are subject to substantial margins of error as indicated, and in no way reflect the official view of the World Bank, its Executive Directors, or the countries they represent. 20 Table 2: Estimates of Governance, Cont'd Voice and Accountability Political Stability Government Effectiveness 2000101 1997198 2000101 1997198 2000101 1997198 Eat 5.E. N Eat. S.E. N Eat. S.E. N Eat. S.E. N Est. S.E. N Est S.E. N PRY PARAGUAY -0.70 0.24 4 -0 42 0.25 4 -0. 87 0.32 4 -0.57 0.32 3 P -2W 02 IS -1.10 0.33 3 PER PERU 0,~~~~~~~(L15 0 23 4 -0.69 0.25 4 ,-0.23 0.24 8 -0.53 0.25 8 -0 36 019 1'A 0.17 0.24 6 PHL PHILIPPINES 0.63 -0,$\ 0.63 0.5 5 .021 022 9 0.27 0.26 5 43 GP P 0.13 0.25 6 POL POLAND -8 1.12 0.19 6 ;089 0.22' 9g 0.84 0.22 7 .47 0.17 1 1 0.67 0.21 8 PRT PORTUGAL 1 42 0 24 5 1.48 0.25 5 1lA1 0 23 7 1.39 0.25 6 :-OJV0 0 20 7 1.15 0.23 7 PPR PUERTORICO . . -0 83 0.49 1 0.76 0.53 1 '-4.0.43 1 1.26 0.52 1 OAT QATAR -054 0 25 3 -0.76 0.25 3 1 40 0 28 4 1.38 0.30 3 0.82 024 4 0.48 0.32 3 ROM ROMANIA os 50.CLO8 0.29 OAR9 4 -00D8 0 28 1 0.02 0.24 4 -0 54 0.19 9 -0.57 0.26 4 RUS RUSSIA -0 35 018 8 -0.19 0.19 6 -0.41 0 22 9 -0.69 0.22 7 -0 57 01? 1 -0.59 0.21 8 RWA RWANDA -1.42 (036 2 -1.17 0.36 1 -11.18 0863 1 . STP SAO TOME AND PRINCIPE 1 00 039 1 0.82 0.36 1 . . -, .., . SAUJ SAUDI ARABIA -1 07 0,24 4 -1.10 0.25 3 10,51 0.25 5 0.24 0.28 4 04 022 5: -0.35 0.29 4 SEN SENEGAL 0.12 0 24 5 -0.29 0.25 4 -0868 0.31 4 -0.87 0.32 3 '016s 0.21 .1 0.05 0.29 4 SLE SIERRA LEONE -1.35 027 3 -1.62 0.29 2 :-I 26 0 51 2 -1.52 0.41 1 '1.80 0 40 tt 0.01 0.77 1 SGP SINGAPORE 0 11 0.24 6 a.13 0.25 6 1.44 0.23 8 1.39 0.25 6 Vt.# 0.19 P. 2.08 0.23 7 SVK SLOVAK REPUBLIC 0 99 016 7 0.52 0.19 5 082 0.25 7 0.65 0.23 6 0.2 , U. - -0.03 0.22 6 SVN SLOVENIA 17 01 7 103 OR 4 07 24 7 1.09 0.24 4 9* J 4 0.57 0.26 4 SLB SOLOMON ISLANDS 018 0.39 1 1.17 0.36 1 - - . 5DM SOMALIA -1.45 o2i 3 -1.69 0.29 2 -1 55 0 61 2 -1.71 0.41 1 -234 0 40 2 -1.70 0177 1 ZAP SOUTH AFRICA 1.1? on2 7 0.99 0.25 5 007 0 22 9 -0.53 0.24 7 0 28 0.18 10 -0.01 0.21 8 ESP SPAIN 1 15 0 23 7 1.36 0.25 5 1 01 022 9 0.58 0.25 6 1 157 019 9 1.60 0.23 7 LKA SRIL.ANKA -2 024 42-0.16 0.25 3 -1 63 0127 5 -1.63 0.30 3 K.,04 0.22 5 -0.61 0.32 3 SON SUDAN 12 024 ~4 -1.50 0.25 3 -2 01 0 28 4 -1173 0.34 2 -1 34 0 24 4- -1170 0.37 2 SUR SURINAME 0.63 028 2 0.28 0.29 2 012 0.85 1I -0.20 0.41 1 OS .3 1 -0.15 0.77 1 SWZ SWAZILAND -093 0.39 1 -0.78 0.36 1 -- -1.69 0.81 1 .~ .-0.47 0.33 2 SWE SWEDEN 1.65 023 7 1.60 0.25 4 l 38 0.22 '9 1.41 0.26 5 1'51 0.19 9 1.57 0.25 6 CHE SWITZERLAND 1.5 .0 1.68 0.25 5 1861 023 a 1.69 0.25 6 1.93 019 8 1.99 0.23 7 SYR SYRIA .140 0.25 3 -1.36 0.25 3 -0 28 0 28 4 0.08 0.30 3 -0 81 0 24 4 -1.18 0.32 3 TWN TAIWAN 0893 0.23 6 0171 0.25 5 070 0.22 8 0.94 0.26 5 0 91 019 9 1.29 0.25 6 TJK TAJIKISTAN -0 68 0.19 3 -1.13 0.23 2 -1.77 042 2 -1.86 0.33 2 -1 31 0 28 3 -1.42 0.34 2 TZA TANZANIA -0 07 024 5 .-0.28 0.25 4 -0 34 0 27 5 0.57 0.27 5 -0 43 023 5 -0.49 0.22 6 THA TH-AILAND 0 37 0.24 8 0.22 0.25 6 0.21 0 23 8 0.25 0.25 6 010 019 9~ 0.01 0.23 7 TGO TOGO .10 0. 28 2 -1.05 0.29 3 -0862 085 1 -0.91 0.37 2 -1.32 0.53 1- -0.37 osi1 2 770 TRINIDAD AND TOBAGO 0861 0 25 4 0.95 0.25 3 Q 27 0 32 4 0.32 0.34 2 0.82 0 25 4 0.52 0.37 2 TUN TUNISIA -0861 0.25 4 -0.59 0.25 3 Of2 0,27 5 0.66 0.28 4 .130 0.23 6 0.63 0.24 5 TUR TURKEY -0.55 023 7 -0.88 0.25 5 -0 75 0.22 9 -0.94 0.25 6 -0 15 018 10: -0.41 0.23 7 TKM TURKMENISTAN -1.42 0L19 2 -1.45 0.23 2 0.11 0,49 1 0.00 0.33 2 -1.23 032 2 -1.25 0.34 2 UGA UGANDA .0179 0 24 5 -0.52 0.25 4 -1 31 0.27 5 -0.98 0.27 5 -0 32 0.23 5 -0.25 0.23 6 1UKR UKRAINE -0.31 0.18 8 -0.05 0.19 5 -0 59 025 7 -0.24 0.22 7 K.0 75 0i 19 a -0.89 0.21 7 ARE UNITED ARAB EMIRATES -0 51 0,25 3 -0.54 0.25 3 1 09 0.29 3 0.82 0.30 3 0.50 0.26 3 0.14 0.32 3 GBR UNITED KINGDOM 1 46 025 7 1.51 0.25 5 110 022 9 0.92 0.25 6 1.77 0 18 10 1.97 on2 7 USA UNITED STATES 1 2' 0.24 8 1.52 0.25 6 118 0.23 8 1.10 0.25 6 1 58 0 19 9 1.37 0.23 7 URY URUGUAY 1 08 0.24 8 0.77 0.25 3 12D5 028 6 0.35 0.30 3 08% 02 I 7 0.62 0.32 3 UZB UZBEKISTAN -118a 018 5 -1.28 0.21 4 .1 I? 0 29 4 -0.33 0.27 4 4(88 0 22 8 -1.30 0.25 4 yEN VENEZUELA 4L-0.4 02 0.15 0.25 5 -0 33 0 23 7 -0.25 0.25 6 -0.311 0. 19 8 -0.85 0.23 7 VNM VIETNAM .1.29 0L24 4--1.4S 0.25 4 044 024 8 0.65 0.26 5 -0.30 0 20 6 -0.30 0.26 5 WTB WESTEBANK 0 00 1200 <1f -0.13 0.86 1 -0 39 0 90 1 0.11 0.66 1 0.68 0.26 6 0.44. 0.30 3: 0.13 0.21 5 MOZ MOZAMBIQUE -0.6* 0 54 2 -0.23 0.33 4 4t0'?7 3:: -105 0.36 4 020: 0.89 2 -0.53 0.31 3 MMR MYANMAR -4.1 037 3 -1.01 0.40 3 -1.02 C' 24 5 -.084 0.28 4 -tlO 0.26 4 -1.10 0.27 3 NAM NAMIBIA 0.08, 0.63. 3 0.27 0.33 3 124 O2. 5- 0.95 0.31 4 1,26 0.28 4 0.38 0.24 3 NPL NEPAL -0.A1 -.0.8L4: I -0.36 0.60 1 -0856 0 43 2 -0.56 0.60 1 -0.31 0.44. 1 NLD NETHERLANDS I SO 0 29 -1.14 0.23 5 157 *1a *, 1.58 0.22 7 2.09 0.20-. 0 7.03 0.20 7 NZL NEWZEALAND l>3 020 S 1.20 0.23 5 . 4.114 1;019 7 1.62 0.23 6 2.09 . l 0.1 2.07 0.20 6 NIC NICARAGUA 41.jd 0.4t4-- -0.10 0.52 2 47: 0M:0.24k 7 -0.73 0.28 4 -0.8 0.21~ 0 -0.64 0.25 3 NER NIGER -0.30 0.54 2- -0.52 0.52 2 -1.117 O-37' 3- -114 0.44 2 -1 0,0.8 .1 -1.57 0.75 1 NGA NIGERIA -0.39 02 8 -0.35 0.29 5 -4.11 0.2 8:, -1.10 0.25 6 -10 .0 7 -0.95 0.20 6 NOR NORWAY UT73 020~ S 0.93 0.23 5 4.0 09 :8. 1.63 0.22 7 - .78 0.22 7 1.69 0.20 7 OMN OMAN 0>8 1,7 0.30 0-40 3 1>00. 0 24 5 1.08 0.28 4 0r.44 :0.26 4t 0.48 0.27 3 PAK PAKISTAN 430- 0.4- 65 -0.20 0.27 4 '0.74 .0.22 7. -0.76 0.24 6 -0.70 0.24 6 -0.77 0.22 6 PNPANAMA 0.96 0.3 6 1.00 0.40 3 .012 -0 20 ,81-. -0.39 0.28 4 OA.-.5 01 06 02 PNG PAPUANEWGUINEA -0.58 .7 1 -13 0.62 2 -0.28 0 2' 5 - -0.31 0.31 3 -1,21 0 .25: 4 -0.85 0.30 2 Note: 'Est' refers to the point estimate of governance. "S.E." refers to the standard error. "N" refers to the number of sources in which the country appears. Governance indicators are oriented so that higher values correspond to better outcomes, on a scale from -2.5 to 2.5. These ratings are based on subjective assessments from a variety of sources, are subject to substantial margins of error as indicated, and in no way reflect the official view of the World Bank, its Executive Directors, or the countries they represent. 23 Table 2: Estimates of Governance, Cont'd Regulator Quality Rule of Law Control of Corruption 2000101 1997198 20D0/01 1997198 2000/01 1997198 Eest S.E. N Eet. S.E. N Eet. S.E. N Est. S.E. N Est. S.E. N Est. S.E. N PRY PARAGUAY 4143.. 0.45' :3 0.37 0.52 3 -0.8 ,0.2e 6 -0.70 0~29 4 48 0.2,7 4 5 -0.96 0.27 3 PER 'ERU 0.3 ..3. 0867 0.27 5 -0.5$- 0A.19 1* -0.52 0.22 7 -0.04 0.18 10 -0.20 0.21 6 PHL. PHILIPPINES 0.21, ~029 85 0.57 0.23 5 440.49 OA8 10 -0.08 0.22 7 4X49 .QI? 117, -0.23 0.16 7 PaOL POLAND 0.41 -.20 80 0.56 0.19 7 056-% 0,15, ->1 0.54 0.16 1 0 0.43 1115 1 2 0.49 0.16 9 PRT PORTUGAL 0.91 - 029- 6 .0.89 0.23 6 0.4 0.118~ 8: 1.08 0.21 8 1.21. 0,21 7V 1.22 0.19 8 PRI 'UERTO RICO 0.99,,~ 0.46 1 0.84 0.52 1 i0.9 -487, 1 0.77 0.50 1 1,1$ 0.4? I1 1.12 0.53 1 OAT DATAR 0.30 (.3?'. -3 0.33 0.40 3 1.00i~ 0 24 6 1.27 0.28 4 0.67 0.5 4` 0.67 0~27 3 ROM ROMANIA W02 0.0: 0.20 0.26 4 402: 017 9 -0.09 0.19 6 A51 0.10 10, -0.46 0.19 6 RUS RUSSIA -1.40. 0260 >8 -0.30 0.19 7 -048?,0.1 11 -0.72 0.16 10 -1.01 0.16 -12 -0.62 0.16 1 0 RWA RWANDA 4073 0.64 1' -1.17 0.60 1 -VIZ 0 43 2 -1.20 0.60 1 0.36 0,44 1 STP SAO TOME AND PRINCIPE . . . .. . .. . .. . SAU SAUDI ARABIA 4011, W.31 3~ -0.15 0.40 3 0.8 .2 0.49 0.26 5 4O35 .25 6 -0.58 0~28 4 SEN SENEGAL .0.38: 0.3 .3 -0.34 0.52 3 4O13, ~0.29 $ -0.10 0.29 4 4038 026, 4 -0.24 0.27 3 SLE SIERRALEONE -1.21: 0.54 2- -1.50 0.52 2 -0.38 "T3 7 -0.91 0.44 2 -o.46 0.39 2 -0.02 0.75 1 SGP SINGAPORE 1.82 0,O27 7 1.24 0.23 6 1 85 OjIS ;9 1.94 0.21 8 2.13 011 10 1.95 0.16 8 SVK SLOVAK REPUBLIC 0.27 02S I. 0.17 0.22 8 0 3f 04.7 3 0.13 0.17 8 0.23 0.17 9 0.03 0.17 7 SVN SLOVENIA 0.52 a0.28 7 0.53 0.26 4 0.89 016, ,9 ,0.83 0.19 6 1.09 0.16 9 1.02 0 21 4 SLB SOLOMON ISLANDS- ... . .. . 1. . . , -.. SOM SOMALIA -1.308 0.54 2 -1.77 0.52 2 -1.29 0.3T S a -1.50 0.44 2 .-t18 0.39 2 -1.05 075 1 ZA07 SOUTH AFRICA 0.0? T2 0.24 0.20 7 .0.06 0.18. 1 IO 0.35 0.21 9 0.35 0.18 10, 0.30 0.17 9 ESP SPAIN 1.08 -0L2V6 0 off 0.23 8 1.12 0.18 - 10 1.03 0.21 8 1.45 0.119 9 1.21 0.19 8 LKA SIRI LANKA OM3 0.34 :4 0.62 0.40 3 -0431 .1> -0.38 0.28 4 .0.00, 0.23 5 -0.12 0.27 3 SON SUDAN 4-AV 0.O? a -0.83 0.52 2 4V4l 0424 $, -1.35 0.31 3 4.24 0,25 4~ -1.02 0.30 2 SUR SURINAME 4081 0.64 2 -0.59 0.52 2 .0.8 043 2 -0.73 0.44 2 0.13 0A6 1, -0.02 0.75 1 SWZ SWAZILAND 0.23 0.64 I 0.29 0.35 2 0.15 0. ' -0.08 0.57 2 .. . 0.01 0.38 1 SWE SWEDEN 1.08: 0.2 6 0.85 0.23 5 tIfi; 048: 10 -1.62 0.22 7 2.21 0.19 9 2.09 0.20 7 CHE SWITZERLAND 1.2 (1,029 5 0.88 0.23 6 1.91 '048 9 2.00 0.21 8. 1.91 0.20 8 2.07 0.19 8 SYR SYRIA 6 .3 -0.92 04 05025 -0.29 0.28 4 -0.83 0.25 4--0.79 0.27 3 TWN TAIWAN 0.$027- 083 0.23 s Ms 0.7 IS 0.8 9, 0.93 0.22 7 0.53~ AM1 10 0.63 0.16 8 TJK TAJIKISTAN -1 8f 3.-1.52 0.27 3 -1.26 022 -4 -1.33 0.22 4 -LOS: 0.24 $ -1 32 0.26 3 TZA TANZANIA 40.2 4~03? 4 0.18 029 5 0.185 0.24 -8' 0.16 0.25 6 .492 0.23 5S -0.92 0.21 5 THA THAILAND 0.80- -0.2?' 7 0.19 023 8 0.44 018 89 0.41 0.21 8 0.046 04? 10 -0.16 0.16 8 TOO TOGO -OSW -0.44 .2 -0.85 0.52 3 -M.0. 043 '3 -0.80 0.37 3 :448 0.05' 1 -0.24 0.48 2 TTO TRINIDAD AND TOEAGO 0.88. DM114 0.72 0.52 2 O,4 0 25 5 0.51 0.31 3 0.49 0,35 4: 0.51 0.30 2 TUN TUNISIA 0*-',7.4~ -043 0.29 4 0.83 - 0.4 0.65 0.27 5 1~B 0.88 0.3 0.02 0.22 4 TUR TURKEY 0.04 I.02 -7 59 0.23 6 -0.16 0.18 10.-0.01 0.21 8 1 0-4OM 0.16 10~ -0.35 0.19 8 TKM TURKMENISTAN 1. 3<03 '3 -1-93 0.27 3 -1.02 0 23 -3 -0.97 0.22 4 -.1.1-2 027 2, -1.29 0.26 3 LGA IJGANDA -020 0.3?, 4, 0.18 0.29 5 .0.56 0 24 -81-0.01 0.25 6 .0.92 0.23. 5 -0.47 0.21 5 L'KR IJKRAINE -105 0.32, 65 -0.72 0.22 8 -0 63 017 ~-*9 -0.71 0.17 9 4090 0.18 9, -0.89 0.17 9 AIRE IJNITED ARAB EMIRATES OM3 .3 >3 0.30 0.40 3 1 1l2 . 08-4 0.77 0.28 4 - 04 028 - 3 -0.03 0.27 3 GBIR IJNITED KINGDOM 1.3Z2- 027 > 1.21 0.23 6 I181 -018 -10 1.69 0.21 8 . 1.0 Oz8IO,10 1 71 0.19 8 LSA IJNITED STATES 1.St027.- 1.14 0.23 6 158 0.18 79 1.25 0.21 8 1.4A 0.9 1.41 0.20 7 LRY IJRUGUAY 0.96 - 0.3,t>> 4- 095 0.40 3 063 021 8 0.27 0.28 4 0.? I 0.2 a 0.43 0.25 4 LZB tJZBEKISTAN -1.1 03 3 4. -1.40 0.27 4 -0.71 0.20 ,---0.87 0.19 6 -0.68 02D 6 -0.96 0.19 5 VEN VIENEZUELA .0.30 0.27 '7 0.09 0.23 6 Ae0.81'O1.14 '9 -0.66 0.21 8 439a 019 a -0.72 0.20 7 VNM VIETINAM -0.0S .- -0.46 0-27 4 0.? O1 7 -0.44 0.24 6 .0.78 0.18 7 -0.33 0.17 6 VVTB WESTBANK 0.3 06-'I -0.16 0.98 1 0.30 16$ I 1.22 0.56 1 0.88 08-1 3&5 WSM WESTERN SAMOA OA104 1 . . 04 0 J2 1 . ,. YEM YEMEN 403D0 '.54- 3 -0.52 0.52 2 -1,12 0,31'.1 4,: -1.01 0.31 3 470 0290 3$ -0.85 0.30 2 YUG YUGOSLAVIA. FEDERAL REPUBL070 ..1; -54 .2 1 44 0.25 3- -0.81 0.20 4 -1.04 ;024 3 -0.99 0.20 5 ZAR C ongo, Dem. Rep. (Zaire) -2.8? .713 -2.34 0.40 3 -2.l9 0.27 -4 -2.15 0.35 3 -1.24' OM3 3 -1.56 0.48 2 ZMB ZAMBIA 0.490.- 37 A4 025 0.29 5 439- .0.24 8 -0.40 0.25 6 43?f 0.23 6: -0.61 0.21 5 ZWE ZIMBABWE .1.68- 0.34 5 -0.34 0.23 6 -0.84' 0.21. 8iS -0.15 0.23 7 -1.08 0.23 ~5 -0.32 0.18 7 Note: "Est' refers to the point estimate of governance. "S.E" refers to the standard error. "N" refers to the number of sources in which the country appears. Governance indicators are oriented so that higher values correspond to better outcomes, on a scale from -2.5 to 2.5. These ratings are based on subjective assessments from a variety of sources, are subject to substantial margins of error as indicated, and in no way reflect the official view of the World Bank, its Executive Directors, or the countries they represent. 24 Table 3: Summary of Differences Between 2000101 and 1997/98 Indicator Number of Countries Number of Sources Correlation between periods 2000/01 1997/98 2000/01 1997/98 Voice and Accountability 173 173 8 6 0.960 Political Instability 161 155 8 6 0.905 Government Effectiveness 159 156 10 8 0.921 Regulatory Quality 167 166 8 8 0.923 Rule of Law 169 166 10 10 0.966 Control of Corruption 160 155 13 11 0.952 25 Table 4: Assigning Countries to Governance Categories Countries with 90% Confidence Interval Entirely Within: Bottom Third Top Third Voice Number 31 42 % of Total 18% 24% Political Stability Number 22 30 % of Total 14% 19% Government Effectiveness Number 22 33 % of Total 14% 21% Regulatory Quality Number 22 23 % of Total 13% 14% Rule of Law Number 20 38 % of Total 12% 22% Control of Corruption Number 20 39 % of Total 13% 24% 26 Table 5: Parameter Estimates VA VA PS PS GE GE RO RO RL RL cc cc 2000101 1997198 2000/01 1997198 2000,101 1997198 2000/01 11997/98 2000/0 1997198 2000/01 1997(98 Intercept ) CUD 0.550 .. 0.666 .. 0.515 .. . . 0.493 .. 0.366 DRI 0.690 0.370 0.434 0.539 0.759 0.804 0.601 0.668 0.488 0.539 EIU 0.369 0.370 0.511 0.520 0.404 0.432 . .. 0.427 0.442 0.297 0.312 FRH 0.542 0.520 .. . .. . ... ... HWJ .. . . . . 0.482 0.468 0.444 0.466 PRS 0.606 0.605 0.723 0.606 0.570 0.789 0.578 0.608 0.632 0.606 0.464 0.506 WBS 0.344 0.448 0.429 0.448 0.533 0.473 0.494 0.563 0.542 0.354 0.534 0.465 BPS .. . . . . . . . . 0.797 BRI .. . 0.636 0.532 0.391 0.406 . . 0.336 0.404 0.477 0.483 CER .. 0.445 .. . . . . 0.604 .. 0.615 EBR . .. . .. . .. 0.502 0.592 . FHT 0.499 0.532 -. . 0.612 0.624 . .. 0.489 0.566 0.465 0.561 GCS 0.570 0.469 0.372 0.398 0.431 0.470 0.472 0.526 0.567 0.493 GCSA . . .. . .. 0.470 .. 0.431 .. 0.499 .. 0.498 GMS 0.378 0.200 .. 0.115 .. . . 0.198 .. 0.551 0.470 LBO . .. ... ...... 0.364 PWC . .. 0.518 .. 0.567 .. . . 0.683 PRC . . . . . . . . .. 0.286 0.302 WOY 0.378 0.469 0.451 .. 0.277 0.307 0.443 0.392 0.337 0.382 0.160 0.212 Slope (5)) CUD 0.220 .. 0.211 .. 0.218 .. . . 0.241 .. 0.305 DRI . .. 0.226 0.292 0.283 0.239 0.163 0.128 0.225 0.179 0.259 0.221 EIU 0.278 0.292 0.297 0.280 0.241 0.226 . .. 0.268 0.285 0.284 0.309 FRH 0.272 0.280 . . . . . . . HWJ . .. . .. . .. 0.154 0.164 0.272 0.247 PRS 0.246 0.247 0.129 0.247 0.095 0.109 0.152 0.139 0.182 0.220 0.159 0.142 WBS 0.008 0.060 0.082 0.060 0.076 0.097 0.028 0.009 0.083 0.135 0.141 0.150 BRI . . 0.153 0.332 0.145 0.133 . .. 0.232 0.136 0.166 0.173 CER .. . . 0.096 .. . . . . 0.380 .. 0.359 EBR .. . . .. 0.133 0.235 . FHT 0 364 0.332 .. . 0.355 0.396 . . 0.349 0.397 0.333 0.513 GCS . 0.136 0.042 0.126 0.151 0.122 0.259 0.203 0.148 0.160 0.247 GCSA . . .. . .. 0.112 .. 1.105 .. 0.041 .. 0.297 GMS 0.108 .. 0.041 .. 0.041 .. . . 0.038 .. 0.148 0.149 PWC . . . 0.080 .. 0.096 .. . . 0.099 PROC. . . . . . . . . 0.325 0.299 WCY 0.107 0.042 0.283 .. 0.141 0.120 0.181 0.290 0.208 0.156 0.293 0.284 Standard Deviation of Error (a) CUD 0.915 .. 0.801 .. 0.608 .. . . 0.743 .. 0.491 DRI . . 0.569 0.519 0.482 0.58 0.521 0.608 0.398 0.583 0.537 0.618 EIU 0 507 0.519 0.402 0.386 0.366 0.396 . .. 0.567 0.445 0.426 0.322 FRH 0.419 0.386 .. . .. . .. ... HWJ .. . . . . 0.827 0.754 0.606 0.751 PRS 0.409 0.495 0.866 0.495 0.628 1.222 1.007 1.032 0.759 0.656 0.849 1.129 WBS 16.832 1.720 2.068 1.720 0.982 0.685 3.495 4.576 0.929 0.681 0.554 0.636 BPS .... .. ...... 0.568 BRI 0.613 0.301 0.565 0.707 . .. 0.475 0.687 11.2911 1.226 CER .. 1.395 .. . . . . 0.303 .. 0.356 EBR . .. . .. 1.041 0.353 ... FHT 0.223 0.301 . .. 0.476 0.440 . .. 0.322 0.454 0.338 0.511 GOS .. 1.086 2.735 0.536 0.600 0.810 0.382 0.410 0.594 0.502 0.457 GCSA .. . . . . 0.450 .. 0.433 .. 1.695 .. 0.406 GMIS 1.047 .. 2.402 .. 2.680 .. . . 2.533 .. 0.895 0.709 PWC . . . . 0.743 .. 0.860 .. . . 0.659 PRO . . . . . . . . .. 0.3S1 0.270 WCY 1.060 2.735 0.590 .. 0.675 0.906 0.547 0.415 0.521 0.580 0.516 0.498 27 Table 6: Variance Decomposition for Changes in Governance Estimates (1) (2) (3) (4) Standard Deviation of Fraction of Variance of Changes in Fraction of Variance of Changes in Fraction of Variance of Changes Changes in Estimated Estimated Governance Due to Balanced Indicator Due to Changes in in Estimated Governance Due to Governance Chanqes in Balanced Indicator Underlvinq Sources Chanqes in Underlyinq Sources Voice and Accountability 0.09 0.95 0.97 0.92 Political Stability 0.12 1.05 0.82 0.86 Government Effectiveness 0.12 0.95 0.88 0.84 Regulatory Quality 0.11 0.80 0.95 0.76 Rule of Law 0.08 0.67 0.85 0.57 Control of Corruption 0.12 0.46 0.92 0.43 28 Figure 1: Estimates of Governance 2.5 Voice and Accountability 2000/01 (1997/98 Values Indicated as Dots) 2 1.5 - -0.5 -1. -15 - -2 -2.5 C ~ 0 D E m N 0 j 0 X z D 0 i M W ~~ < 0 z a 6 , z , M < 0z < Of Z Note: This graph shows es timates of the indicated dimension of governance (on the vertical axis) for all countries for which data is available (on the horizontal axis) for 2000/01. The vertical bars show the statistically-likely range of values of governance for each country, with the midpoint of each bar corresponding to the best single estimate. The length of these bars varies with the amount of information available for each country, and the extent to which information from different sources correspond with eachother. Estimates of governance for 1997/98 are indicated as dots. Selected countries are indicated on the horizontal axis. As emphasized in the text, the ranking of countries along the horizontal axis is subject to significant margins of error, and this ordering in no way reflects the official view of the World Bank, its Executive Directors, or the countries they represent. 29 Figure 1, Cont'd: Estimates of Governance 2.5 Political Stability 2000/01 (1997198 Values Indicated as Dots, All Sources) 2 - 1.5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~10 U~~~~~~~~~~~~~~~~~~~~~~~~T _T _T At XT TT 0.5 0-1 -1.5I I -2. vE. z E z zM EE OX N < - > * C Z Z - U , , ,) a ,,, N o Ul n w w al 7 o< U) DOW 0 0 U) of CX l )2- °8N9 a: 0 < GoD z IM z z U U) W )~m0 2 Note: This graph shows estimates of the indicated dimension of governance (on the vertical axis) for all countries for which data is available (on the horizontal axis) for 2000/01. The vertical bars show the statistically-likely range of values of governance for each country, with the midpoint of each bar corresponding to the best single estimate. The length of these bars varies with the amount of information available for each country, and the extent to which information from different sources correspond with eachother. Estimates of governance for 1997/98 are indicated as dots. Selected countries are indicated on the horizontal axis. As emphasized in the text, the ranking of countries along the horizontal axis is subject to significant margins of error, and this ordering in no way reflects the official view of the World Bank, its Executive Directors, or the countries they represent. 30 Figure 1, Cont'd: Estimates of Governance 2.5 2_ - Government Effectiveness 2000101 (1997-98 Values Indicated as Dots) 2 - 1.5 -1 -1.5 -2.5 -2.5 M. Z P: 3 ui CO Z 0 0 0 N < - >0 Z X v< 0 x X z 0 -J - : cw 0 O 2 = 2 W 0 0 Z -J Z m W~~~~~ Lu < ) - cn 2Z M- Lu><0 - 0 M 0 2 N Y z -ZD -00 u - < 0 00< Note: This graph shows estimates of the indicated dimension of governance (on the vertical axis) for all countries for which data is available (on the horizontal axis) for 2000/01. The vertical bars show the statistically-likely range of values of governance for each country, with the midpoint of each bar corresponding to the best single estimate. The length of these bars varies with the amount of information available for each country, and the extent to which information from different sources correspond with eachother. Estimates of governance for 1997/98 are indicated as dots. Selected countries are indicated on the horizontal axis. As emphasized in the text, the ranking of countries along the horizontal axis is subject to significant margins of error, and this ordering in no way reflects the official view of the World Bank, its Executive Directors, or the countries they represent. 31 Figure 1, Cont'd: Estimates of Governance 2.5 Regulatory Quality 2000-01 (1997-98 Values Indicated as Dots) 2- 1.5 1 lil~~~~~~~~~~~TT t- LT T T.TI .T 0.5- 1 ' I ggLlel iX 111 -1 HJ 0 -i~~~~~~~~~~~~~~~~~~~~I -2 -2.5 <0 _ - CX1 W~ <~ C . OT < D H < I- a 0H E ZU N w HO> H > 0 0 DI 00 u a > 0w T a F 0 ; e- 0 >- F > °Z°t z ° (D alot mC 2 -D z eco OXz < Note: This graph shows estimates of the indicated dimension of governance (on the vertical axis) for all countries for which data is available (on the horizontal axis) for 2000/01. The vertical bars show the statistically-likely range of values of governance for each country, with the midpoint of each bar corresponding to the best single estimate. The length of these bars varies with the amount of information available for each country, and the exte nt to which information from different sources correspond with eachother. Estimates of governance for 1997/98 are indicated as dots. Selected countries are indicated on the horizontal axis. As emphasized in the text, the ranking of countries along the horizontal axis is subject to significant margins of error, and this ordering in no way reflects the official view of the World Bank, its Executive Directors, or the countries they represent. 32 Figure 1, Cont'd: Estimates of Governance 2.5 Rule of Law 2000/01 (1997/98 Values Indicated as Dots) 2- 1.5- 0.5 -0.5- -1.5- -2- -2.5 wN )< < Z () C .Z ZW L OW v Z I 2 B Z m 0 m ° 2 X I z m ° < Z E m g 2 B CD C) 0z ~~g~w I O > f t > z ED 6 0 O 1 C ZO z z A- CDwuzC z~c ,- wD LU z-2~ < A- > e >- 0 0 ~~~~~~~~~~~~~~~0 CER = CER m CER O EB ET EBR EB FHT FHT _ FHT . . ...._ _) GCS GCS GCS GCSA GCSA 0 7 GCSA GM- QS GMS | -i=- LBO LBO _LBO jJL PWRC PRC PWRC WCY wCY WCY Q. IOL _ _ _ _ _ _ _ _ _ _ 4__ _ _ _ o e e 00 0 0 000eoo000oo o oo CUD CUD' CUD DRI DRI 0 EIU - EIU EIU FRH FRH FRH. HWJ HWJ HWJ 5. 0~~~~~~~~~~~~~~~~~~0 PRS T PRSF PRS !Q WBS 0 W WBS_W BPS BPS 0 B_ n BPS 0 l BRO 0 BRO o BRO O CER : CER 5 CERQ EBR EBR EBR = O FHT - .FHT, r H GCS SGC GCSA GCSA GCSA G MS GMS * 0Ei + WCY &GMS LBO LBO FI M I~~~~~~~0I ~ ~ ~ ~ ~~0LBO PWC PWC :I PRC ~ ~~~~~~~3PRCPW WCY WCY PRC WCY Table Al Business Environment Risk Intelligence (BRI) http://www.beri.com BERI S.A. is a private source of analysis and forecasts of the business environment in developed and developinq countries. The firm was founded in 1966 and is headquartered in Geneva, Switzerland. BERI has two services that include variables of interest for the purpose of this paper: The Business Risk Service, and The FORELEND or Lender Risk Rating. Both services are supervised by Dr. F.T. Haner, founder and senior editor. A number of analysts review various data sources and produce initial draft reports, relying on an intemational network of sources for intelligence in the field. BERI convenes two permanent panels of about 105 experts from all over the world. These panels provide country ratings and qualitative observations on the basis of these initial reports. One panel assesses political conditions, and the other offers perspectives on the business operating environment. These ratings are constructed using the Delphi method, in which panelists are also supplied with the ratings they produced in previous assessments as well as the panel average score for BRS monitors 50 countries three times per year, assessing 57 criteria separated into three indices. The Political Risk Index (PRI) focuses on sociopolitical conditions in a country. Diplomats and political scientists rate the present condition of eight causes and two symptoms of political risk, using a scale from 7 (no problem) to 0 (prohibitive problem). The Operation Risk Index (ORI) identifies major bottlenecks for business development, rating 15 criteria on a scale of 0 (unacceptable conditions) to 4 (superior conditions). The R factor assesses a country's willingness to allow foreign companies to convert and repatriate profits and to import components, equipment and raw materials. It is composed of 4 sub-indices, one of which assesses the quality of legal framework in terms of statutorv laws and actual practice. In the table below we list the variables included in each of the governance indicators. We use BERI's data for the first quarter of 2001. Business Environment Risk Intelligence: By Aaaregate Governance Indicator Voice and Accountability NA Political Stability Political Risk Index: Internal Causes of Political Risk: Fractionalization of political spectrum and the power of these factions. Political Risk Index: Internal Causes of Political Risk: Fractionalization by language, ethnic and/or religious qroups and the Power of these factions. Political Risk Index: Intemal Causes of Political Risk: Restrictive (coercive) measures required to retain power. Political Risk Index: Intemal Causes of Political Risk: Organization and strength of forces for a radical Political Risk Index: Svmptoms of Political Risk: Societal conflict involvinq demonstrations, strikes, and street Political Risk Index: Symptoms of Political Risk: Instability as perceived by non-constitutional changes, assassinations, and querilla wars. Government Effectiveness Operation Risk Index: Bureaucratic delays Regulatory Quality NA Rule of Law Operation Risk Index: Enforceability of contracts Control of Corruption Political Risk Index: Internal Causes of Political Risk: Mentality, including xenophobia, nationalism, corruption, nepotism, willinaness to compromise. etc. 36 Table A2 Columbia University State Capacity Survey (CUD) Marc Levy of the CIESIN at Columbia University has directed a task force for the implementation of a poll of country experts and the creation of a database measuring state capacity. Data on 109 countries were obtained from assessments completed bV 209 experts durinq 2001. The survey instrument consists of 31 multiple-choice questions and three open-ended questions. The survey includes five broad categories of questions: political context, state legitimacy, human resources and organizations, institutions and overall capacity. In the table below we list the questions included in each of the govemance indicators. Measuring State Capacity Throuah an Expert Survey: By Aaareaate Govemance Indicator Voice and Accountability Question 10. For the most part, is the state seen as legitimately representing its citizens? Political Stability Question 3. Assess the degree to which the decline or collapse of central political authority posed a threat to political stability in this country. Question 5. Assess the degree to which political protest or rebellion posed a threat to political stability in this Question 7. Assess the degree to which ethno-cultural and/or religious conflict posed a threat to political stability in this country. countrv. Question 11. To what extent does the state and/or its allied groups engage in repression of its citizens? Government Effectiveness Question 13. Rate the administrative and technical skills of the country's civil service (occupying middle and hiqher manaqement roles). Question 16. Rate the efficiency of the country's national bureaucracies overall. Question 17. Rate the efficiency of the country's local-level government bureaucracies overall. Question 18. Rate the effectiveness of coordination between the central government and local-level government organizations. Question 21. Rate the state's abilitv to formulate and implement national policv initiatives. Question 22. Rate the state's effectiveness at collecting taxes or other forms of government revenue. Question 23. Does the central government produce a national budget in a timely manner? Question 24. Do local governments produce budgets in a timely manner? Question 27. Rate the state's ability to monitor socioeconomic trends, activities, and conditions within its borders Question 28. Rate the state's ability to create, deliver, and maintain vital national infrastructure. Question 29. Rate the state's ability to respond effectively to domestic economic problems. Question 30. Rate the effectiveness of the state's coercive responses to domestic political crises, including political protest, rebellion, and ethno-cultural or religious conflict. Here effectiveness refers to the states ability to contain or control the crisis. Question 31. Rate the state's ability to respond effectively to natural disasters. Regulatory Quality NA Rule of Law Question 20. In carrying out internal security tasks, to what extent does the state rely on tactics commonly considered illegitimate in the international community? Question 25. Rate the state's adherence to the rule of law, considerinq the country as a whole. Question 26. Is there significant variation in how the rule of law is applied across groups within the country? Control of Corruption Question 15. Rate the severity of corruption within the state 37 Table A3 Standard and Poor's DRI/McGraw-Hill (DRI) http:/leconomics.dri-wefa.com/main.html DRI is an economic consulting and information company which provides data, analysis, forecasts and expert advice to strategic planners, business and financial analysts, and policy makers. It was founded in 1973 as a unit of Standard & Poor's and is based in the United States. In 1996, DRI launched the Country Risk Review (CRR), a quarterly publication providing country rsk assessments to international investors. A first draft of the risk ratings in this publication are produced by country analysts, who then submit their preliminary assessment to regional review committees charged with analyzing and challenging these assessment. The global risk service committee evaluates the reviewed assessments to ensure quality and cross-country consistency. The country analvsts then Produce the final countrv risk review. The CRR assesses the relationship between country risk and its effects on the profitability of investments. For each country, DRI identifies a number of "potential sources of risk", specifies measurable "rsk events", measures how probable those risk events are, and assesses the severity of impact that each outcome would have. Based on these considerations, DRI produces a risk score for each country. The CRR identifies a total of 33 "immediate risk events" and 18 "secondary risk events" for 111 developed and developing countries. Immediate risk events are classified into policy risks (tax, and non-tax), and outcome risks (price, and non-price). Secondary risk events are classified into domestic political risks, external political risks, and economic risks. These risk events are described in below. For each risk event, DRI produces a short run and a long nun risk rating. These ratings provide subjective estimates of the likelihood that a particular risk event will occur within one and five years respectively. DRI follows a methodology to ensure that the five year forecasts are consistent with the short-term forecasts. Although these indicators nominally measure the likelihood of future changes in governance concepts, in practice the lonq-run ratinqs provide qood measures of the current levels of qovernance. In the table below we list the variables included in each of the governance indicators. Variable definitions consist of risk events. The actual ratings provide an estimated probability of these events happening. In this paper, we use data for the fourth quarter of 2000. Standard and Poor's DRI/McGraw-Hill: by Aaprecate Governance Indicator Voice and Accountability NA Political Stability Domestic Political Risks: Military Coup Risk: A military coup d'etat (or a series of such events) that reduces the GDP qrowth rate by 2% durinq any 12-month period. Domestic Political Risks: Major Insurgency/Rebellion: An increase in scope or intensity of one or more insuraencies/rebellions that reduces the GDP qrowth rate by 3% durinq anv 12-month period. Domestic Political Risks: Political Terrorism: An increase in scope or intensity of terrorism that reduces the GDP growth rate by 1% during any 12-month period. Domestic Political Risks: Political Assassination: A political assassination (or a series of such events) that reduces the GDP qrowth rate by 1% durina any 12-month period. Domestic Political Risks: Civil War: An increase in scope or intensity of one or more civil wars that reduces the GDP qrowth rate bv 4% during anv 12-month period. Domestic Political Risks: Major Urban Riot: An increase in scope, intensity, or frequency of rioting that reduces the GDP growth rate by 1% durinq any 12-month period. 38 Table A3, Cont'd Standard and Poor's DRI/McGraw-Hill: by Aggregate Governance Indicator (Cont'd) Government Effectiveness Domestic Political Risk Government Instability: An increase in government personnel turnover rate at senior levels that reduces the GDP crowth rate bv 2% durinq anv 12-month period. Domestic Political Risk: Government Ineffectiveness: A decline in government personnel quality at any level that reduces the GDP qrowth rate bv 1% durinq anv 12-month period. Domestic Political Risk: Insbtutional Failure: A deterioration of government capacity to cope with national problems as a result of institutional riqiditv or qridlock that reduces the GDP qrowth rate bv 1% durinq anv 12- Reaulatorv Quality Policies Non-Tax: Regulations -- Exports: A 2% reduction in export volume as a result of a worsening in export regulations or restrictions (such as export limits) during any 12-month period, with respect to the level at the time of the assessment. Policies Non-Tax: Regulations -- Imports: A 2% reduction in import volume as a result of a worsening in import regulations or restrictions (such as import quotas) during any 12-month period, with respect to the level at the time of the assessment. Policies Non-Tax: Regulations -- Other Business: An increase in other regulatory burdens, with respect to the level at the time of the assessment, that reduces total aqqreqate investment in real LCU terms by 10% Policies Non-Tax: Ownership of Business by Non-Residents: A 1-point increase on a scale from "0" to "10" in legal restrictions on ownership of business by non-residents during any 12-month period. Policies Non-Tax: Ownership of Equities by Non-Residents: A 1-point increase on a scale from "0" to `10" in leqal restrictions on ownershin of equities bv non-residents durinq any 12-month period. Rule of Law Outcomes Non-Pnce. Losses and Costs of Crime: A 1-point increase on a scale from "0" to "10" in crime during anv 12-month period. Domestic Political Risk: Kidnapping of Foreigners: An increase in scope, intensity, or frequency of kidnapping of foreiqners that reduces the GDP qrowth rate by 1% durinq anv 12-month period. Policies Non-Tax: EnforceaUility of Government Contracts: A 1 point decline on a scale from "0" to "10" in the enforceability of contracts during any 12-month period. Policies Non-Tax: Policies Non-Tax: Enforceability of Prvate Contracts: A 1-point decline on a scale from "0" to "10" in the leqal enfor-:ea-'W..l tv of contracts durinq anv 12-month period. Control of Corruption Risk Event Outcome nonr-prce: Losses and Costs of Corruption: A 1-point increase on a scale from "0" to"10" in corruption durinq anv 12-morith Deriod. 39 Table A4 European Bank for Reconstruction and Development (EBR) http://www.ebrd.orq The EBRD is an international organization which supports the transition towards open market-oriented economies and promotes private and entrepreneurial initiative in the countries of Central and Eastem Europe and the Commonwealth of Independent States (CIS). The EBRD is based in London. The EBRD publishes an annual Transition Report, which includes a number of governance variables in its Transition Indicators and Survey of Legal Reforms. The Transition Report presents eight 'Transition Indicators" representing "cumulative progress in the movement from a centrally planned economy to a market economy" for 26 transition economies. The subjective indicators are based on a checklist of various objective measures and reflect the views of EBRD staff. Beginning In 1998, the EBRD has conducted a survey of local public officials, private firms, academics, lawyers, and other experts, in order to assess the progress made in financial legal reform in transition economies. The survey considered two areas of financial legal reform: banking and securities activities. For each area, two indices describing the extensiveness and effectiveness of the financial legal framework were developed, for a total of four ratings. The 'extensiveness" ratings measure how closely legal rules affecting investment follow international standards. "Effectiveness" reflects how clear, accessible and adequately-supported the legal rules are. Both are intended to provide a measure of how conducive the laws of these countries are to fostering investment. This survev covered 26 countries. In this paper we use data from the 2001 Transition Report. In the table below we list the variables included in each of the qovernance indicators. European Bank for Reconstruction and Development bv Aggregate Governance Indicator Voice and Accountability NA Political Stabilitv NA Government Effectiveness NA Requlatory Quality Price liberalisation Trade & foreign exchange system Competition policy Commercial Law Extensiveness Commercial Law Effectiveness Financial Regulations: extensiveness Financial regulations: effectiveness Large-scale privatisation Small-scale privatisation Governance & enterprise restructuring Banking reform & interest rate liberalisation Securities markets & non-bank financial institutions Bancruptcy law Rule of Law NA Control of Corruption NA 40 Table A5 The Economist Intelligence Unit (EIU) http:llwww.eiu.com The Economist Intelligence Unit is a for-profit organization producing analysis and forecasts of the political, economic and business environment in more than 180 countries. The EIU was founded in 1949 and is based in London. In 1997, the EIU launched two quarterly publications which contain some govemance measures: The Country Risk Service, and the Country Forecasts. The assessments in these publications are based on regular contributions from a global network of more than 500 information-gatherers. A panel of regional experts checks the accuracv, consistencv and impartiality of these assessments. Country Risk Service (CRS). The CRS provides intemational investors with risk ratings for seven broad categories of country risk-Political, Economic Policy, Economic Structure, Liquidity, Currency, Sovereign Debt, and Banking Sector-and covers 100 emerging markets. Within each category, subjective assessments are produced for a number of sub-issues. Country Forecasts (CF). The CFs measure the quality or attractiveness of the business environment in 60 countries. The rankings examine ten separate criteria-the political environment, the macroeconomic environment, market opportunities, policy towards free enterprise and competition, policy towards foreign investment, foreign trade and exchange controls, taxes, financing, the labor market and infrastructure-and are designed to reflect the main criteria used by companies to formulate their global business strategies. Each criteria covers a number issues for which two assessments are produced: A historical rating, covering the last 5 vears, and a forecast ratinq coverinq the next five vears. For the purpose of this paper. the most interestinq Combining the CRS and CF Ratings. The Political Risk rating from the CRS and the Political Environment ratings from the CF ratings cover almost identical concepts, although using different rating scales. We convert the CF ratings to the same units as the CRS ratings, and augment the CRS ratings with the CF ratings for the 15 mostly OECD economies appearing in the CF but not the CRS. This results in an indicator covering 115 developed and developinq countries. In the table below we list the variables included in each of the governance indicators. In this paper, we combine data from the 2001 first quarter edition of the Country Risk Service, and from the 2001 historical risk rating from the Countrv Forecast. Economist Intelligence Unit: by Aggregate Governance Indicator Voice and Accountability Political Stability: Change in govemment, orderly transfer Political Effectiveness: Legal system / transparency and fairness Political Stability Political Stability: Armed conflict / War Political Stability: Social unrest Political Stability: Politically motivated violence / Terrorist threat Government Effectiveness Political Effectiveness: Govt. policy / Pro bussiness orientation Political Effectiveness: Govemment efficacy / Institutional effectiveness Political Effectiveness: Bureaucracy / Red tape Regulatory Quality NA Rule of Law Political Effectiveness: Crime Corruption of Bank Officials Control of Corruption Political Effectiveness: Corruption among public officials 41 Table A6 Freedom House (FRH, FNT) http://www.freedomhouse.orq. Freedom House is a non-governmental organization which promotes democratic values around the world. Freedom House was established in 1941 and is headquartered in New York Citv. We rely on data from two Freedom House publications. "Freedom in the World was launched in 1955", and became an annual publication in 1978, and covers 191 countries and 60 related territories. "Nations in Transit" was launched in 1995 and covers 28 post-communist countries. Freedom House develops its assessments using a team of academic advisors, in-house experts, published resources, and local correspondents including human rights activists, journalists, editors and political figures. Freedom House staff also conduct regular fact-finding missions to countries beinq assessed. An academic advisory board provides input to the Droiect in qeneral. Freedom in the World (FRH). This publication evaluates political rights and civil liberties around the world. Freedom House defines political rghts as those freedoms that enable people to participate freely in the political process, and civil liberties as the freedom to develop views, institutions and personal autonomy apart from the state. For all countries, the subjectiveassessments are based on checklists of rights and freedoms. A Freedom House team assigns a rating to each item on the checklist and produces an initial assessment for each country. The team then assess whether the checklists might have missed an important factor for a particular country. The scores are then reviewed to ensure qualitv and consistency across countries, and a final ratinq is produced. Freedom House Nations in Transit (FNT). This publication evaluates the progress in democratic and economic reform in post-communist countries. Country surveys are written by Freedom House staff or consultants and are reviewed by academics and senior Freedom House staff. Each report is divided into nine sections, ranging from the political process to progress in price liberalization. For each section, a preliminary rating is based on a checklist of issues. The academic oversight board establishes the final ratings by consensus followina extensive discussions and debate. which are reviewed bv the Freedom House ratina In the table below we list the variables included in each of the govemance indicators. In this paper we use data from the 2001-2002 edition of Freedom in the World and the 20001 edition of Nations in Transit. Freedom in the World / Nations in Transit: By Aggregate Governance Indicator Voice and Accountability Political Rights Is the head of state and/or head of qovernment or other chief authoritv elected throuph free and fair elections? Are the legislative representatives elected through free and fair elections? Are there fair electoral laws? Are the voters able to endow their freely elected representatives with real power? Do the people have the right to freely organize in different political parties or other competitive political groupings of their choice, and is the system open to the rise and fall of these competing parties or groupings? Is there a significant opposition vote, de facto opposition power, and a realistic possibility for the opposition to increase its support or qain power throuqh elections? Are the people free from domination by the military, foreign powers, totalitarian parties, religious hierarchies, economic oligarchies or any other powerful groups? Do cultural, ethnic, religious and other minority groups have reasonable self-determination, self-government, autonomy or participation through informal consensus in the decision-making process? 42 Table A6, Cont'd Freedom in the World I Nations in Transit: By Aggregate Governance Indicator, Contd Civil Liberties Are there free and independent media, literature and other cultural expressions? (Note: in cases where the media are state controlled but offer pluralistic points of view, the Survey gives the system credit Is there open public discussion and free private discussion? Is there freedom of assembly and demonstration? Is there freedom of political or quasi-political organization? Are citizens equal under the law, with access to an independent, nondiscriminatory judiciary, and are they respected by the security forces? Is there protection from political terror, and from unjustified imprisonment, exile or torture, whether by groups that support or oppose the system, and freedom from war or insurgency situations? Are there free trade unions and peasant organizations or equivalents, and is there effective collective Are there free professional and other private organizations? Are there free businesses or cooperatives? Are there free religious institutions and free private and public religious expressions? Are there personal social freedoms, which include such aspects as gender equality, property rights, freedom of movement, choice of residence, and choice of marriage and size of family? employers, union leaders, bureaucrats or any other type of denigrating obstacle to a share of legitimate economic gains? Is there freedom from extreme government indifference and corruption Freedom of the Press Laws and Practice: Assess whether or not dissent is allowed, if private media are permitted alongside governmental broadcasting, if independent media, in practice, are permitted to express diverse views Political Influence over Media Content: This category reflects political pressure on the content of both privately owned and govemment media, and takes into account the day-to-day conditions in which journalists work, threats from organized crime, or from religious extremists, for example, often generate self-censorship and so negatively affect the media environment Economic influence over Media Content: Economic influence may come from the government or from private entrepreneurs. This reflects competitive pressures in the private sector that distort reportage as well as economic favoritism or reprisals by government for unwanted press coverage Actual Incident of Violations of Press Freedom: Murders, arrests, suspension and other violations create a sense of fear which may discourage objective reporting Nations in Transit Political Process: Deals with elections, referenda, party configuration, conditions for political competition, and popular participation in elections. Civil Society: Highlights the degree to which volunteerism, trade unionism, and professional associations exist, and whether civic organizations are influential Independent Media: Press freedom, public access to a variety of information sources, and the independence of those sources from undue government or other influences. Political Stability NA Government Effectiveness Nations in Transit Government and Administration: Government decentralization, independent and responsibilities or local and regional governments, and legislative and executive transparency are discussed. Regulatory Quality N/A Rule of Law Nations in Transit minorities. Control of Corruption Nations in Transit Corruption '*13 Table A7 World Economic Forum (GCS) http://www.weforum.orq The World Economic Forum (WEF) is an independent, not-for-profit organization bringing together top leaders from business, government, academia and the media to address key economic, social and political issues in partnership. The WEF was founded in 1971 and is headquartered in Geneva, Switzerland. Since 1996, The WEF has sponsored the Global Competitiveness Report, an annual publication produced in collaboration with the Harvard Institute for International Development (HIID). As background for this report, the WEF conducts the Global Competitiveness Survey of about 3000 enterprises in 60 countries. This survey measures the perceptions of business executives about the country in which they operate. The survey asks top managers to rank on a 1 to 7 scale their opinion on issues in eight broad areas: 1) Openness, 2) Government, 3) Finance. 4) Infrastructure, 5) Technoloqy, 6) Manaqement, 7) Labor, and 8) Institutions. In 1997 the WEF sponsored a separate survey of countries in Africa. In the 1997-98 version of the governance indicators, we included this as a separate source since it asked several questions not included in the main GCS. In 2001, the GCS and the World Bank's World Business Enivronment Survey (WBS) collaborated on a common questionnaire. In order not to use the same information twice, we include these countries in the WBS and exclude them from the GCS. In the table below we list the variables included in each of the governance indicators. In this paper, we use data from the 2001 GCS. Global Comoetitiveness Survey: Bv Apprecate Governance Indicator Voice and Accountability NA Political Stability Likelihood of dramatic changes in institutions New government honors commitments of previous govemment Government Effectiveness Competence of public sector personnel Wasteful government expenditure New government honors commitments of previous govemment Manaqement time spent with bureaucracv Public Service vulnerability to political pressure Govemment economic policies are independent of pressure from special interest groups. Regulatory Quality Tax svstem hinders performance Administrative regulations that constrain businesses are minimal. Rule of Law Extent of tax evasion Costs of organized crime for business Police effectiveness in safeguarding personal security Independence of the judiciary from interference by the govemment and/or parties to the dispute Private business has recourse to independent and impartial courts for challenging the legality of government Financial Assets and wealth are well protected Private businesses are morel likelv to settle disputes outside courts. Control of Corruption Additional payments: bureaucracv Additional payments: judiciary 44 Table A8 Gallup Millennium Survey hftp://www.qallup-intemational.com Gallup International was founded in May 1947, is registered in Zurich, Switzerland, and has 55 members around the world qovemed by the same Code of Statutes to ensure technical competence and quality standards. The Gallup International Millennium Survey (GMS) polled 57,000 adults in 60 different countries of the world between August and October, 1999. The survey covered a wide range of topics of an ethical, political and religious nature, focusing specifically on issues related to democracy, the United Nations, human rights, women's rights, environment, religion, crime and basic values. Slnce this source asks several questions which also appeared in the 1997 Gallup 50th Anniversary Survey which we used in the 1997-98 version of the aqgreqate qovernance indicators, we treat this source as the continuation of the 50th Anniversary Survey. In the table below we list the variables included in each of the governance indicators. Gallup Millenium Survey: By Aagrecate Governance Indicator Voice and Accountability Percent who believe the country is govemed by the will of the people Percent who belive elections are free and fair Percent who believe the govemment is accountable Freedom of speech: 1 (fully respected) - 3 (not respected) Political Stability NA Government Effectiveness Percent who believe the government is efficient Regulatory Quality NA Rule of Law Concern with level of crime Control of Corruption Percent who believe the government is corrupt 45 Table A9 Heritage Foundation/Wall Street Journal (HWJ) http://www.heritaqe.ora The Heritage Foundation is a research and educational institute whose mission is to formulate and promote conservative Public policies. The Heritaqe Foundation was establishec in '973. In 1995 the Heritage Foundation, in partnership with the Wall street Joornal, launched its annual Index of Economic Freedom. This index covers 161 countries and measures economic freedoms and prospects for growth in the global economy. The index is designed for cross country research and is intended to assist international investors and aid donors in the allocation of their rescurces. Tns index is based on a detailed assessment of 10 different factors, including foreign investment cones, taxes, tariffs, banking regulations, monetary policy, and the black market. For some of these, assessments are mnechanically based on objective data, while others are qenerated as subiective ratinqs based on a pre-snecified chrecKlist. In the table below we list the variables included in each of the governance indl.ators. In this paper, we use Heritage data for 2001. Heritage Foundation I Wall Street Journal: Index of Economic Freedom By Aggreagate Governance Indicator Voice and Accountability NA Political Stabilitv NA Government Effectiveness NA Regulatory Quality Regulation Government Intervention Wage/Prices Trade Foreign investment Banking Rule of Law Black market Property Rights Control of Corruption NA 46 Table Al 0 Latinobarometro (LBO) http://www.latinobarometro.cl/Enqlish/Entradai.htm Latinobarometro is a public opinion survey representing the opinions, attitudes, behaviour and values of citizens of the countries in which it is conducted. The survey began being applied regularly in 8 countries of the region in 1995, and in 17 countries beginning in 1997. Latinobarometro conducts an annual survey, using representative samples and an identical questionnaire in each country. It asks questions in in the following areas: Economy and International Trade, Integration and Regional Trading Blocks, -Democracy, Politics and Institutions, Social Policies, Civic Culture, Social Capital and Social Fraud, The Environment, Current Issues. In the table below we list the var.ables included in each of the govemance indicators. We use data from the 2001 LBO. Latinobarometro: by Aggrecate Governance Indicator Voice and Accountability NA Political Stability NA Government Effectiveness NA Requlatory Quality NA Rule of Law Feeling of personal safety Equal oDportunities to access iustice Equality before the law: Control of Corrupztion Percent of public officials viewed to be corrupt. 47 Table All PriceWaterhouseCoopers (PWC) http://www.opacitvindex.com/ PricewaterhouseCoopers (PWC) is a U.S.-based professional services firm. It has set up an "Endowment for Transparency and Sustainability" aimed at supporting research efforts world-wide that shed light on two related topics of global importance: transparency in business and government, and sustainable economic development. Using a team of economists, survey professionals, analysts, and distinguished advisors, it has constructed an "Opacitv Index" measurina the lack of transparency in 35 countries. Opacity is defined as "the lack of clear, accurate, formal, easily discernible, and widely accepted practices" in the following areas: corruption in govemment bureaucracy, laws governing contracts or property rights, economic policies, accounting standards, and business regulation. The index was constructed based on responses to a survey of chief financial officers of medium- and large firms, equity analysts, bankers, and PWC employees resident in each country surveyed. The survey was conducted in 35 industrial and major developing countries durinq the second and third quarter of 2000. In the table below we list the variables included in each of the governance indicators. Opacity Factor: by Aggregate Govemance Indicator Voice and Accountabilitv NA Political Stability NA Government Effectiveness Economic Regulatory Quality Requlation Rule of Law NA Control of Corruption Corruption 48 Table A12 Political Economic Risk Consultancy (PRC) http://www.asiarisk.coml The Political and Economic Risk Consultancy providesstrategic information and analysis for companies doing business in the countries in East and Southeast Asia. PERC was founded in 1976 and is headquartered in Hong Kong. PERC conducts a variety of surveys of expatriate business managers in 12 countries in East Asia. Here we draw on one such survey 'Corruption in Asia in 2001", published in their Asian Intelligence Issue #579 March 7, 2001. In this survey, foreign managers working within the East Asia region were questioned about their perception of corruption, the quality of the legal system, and the professionalism and reliability of the police and judiciary. We have obtained their data on corruption for 12 countries, based over 1000 responses. With respect to corruption, respondents were asked "To what extent does corruption exist in a way that detracts from the business environment for foreign companies?". In the table below we list the variables included in each of the governance indicators. Political Economic Risk Consultancy: By Aaareaate Governance Indicator Voice and Accountability NA Political Stability NA Government Effectiveness NA Regulatory Quality NA Rule of Law NA Control of Corruption Corruption 49 Table A13 Political Risk Services (PRS) http:/lwww.prsqroup.com The PRS group is an affiliate of Investment Business with Knowledge (IBC), a United States-based corporation providing up-to-date country information for international business. PRS was founded in 1980 and is headquartered in Syracuse, New York. Since 1982, PRS produces the Intemational Country Risk Guide (ICRG) which provides assessments of a political, economic and financial risks in a large number of developed and developing countries. These assessments are based on the analysis of a worldwide network of experts, and is subject to a peer review process at subject and regional levels to ensure the coherence and comparability across countries. The ICRG assesses three major categories of risk: political (with 12 components), financial (5 components) and economic (6 components). We use components of the Political Risk Index, which report subjective assessments of the factors influencing the business environment in a particular country. In the table below we list the variables included in each of the governance indicators. In this paper we use data from the August 2001 edition of the International Country Risk Guide, with the exception of the "Investment Profile" question for which we use data from the March 2001 version due to an apparent change in units in April of 2001 for this variable. Political Risk Services by Aaarecate Governance Indicator Voice and Accountability Military in Politics The military are not elected by anyone, so their participation in government, either direct or indirect, reduces accountability and therefore represents a risk. The threat of military intervention might lead as well to an anticipated potentially inefficient change in policy or even in government. It also works as an indication that the qovernment is unable to function effectivelv and that the countrv has an uneasy environment Democratic Accountability. Quantifies how responsive government is to its people, on the basis that the less response there is the more likely is that the government will fall, peacefully or violently. It includes not only if free and fair elections are in place, but also how likely is the qovernment to remain in power or remain popular. Political Stability Infernal Conflict. Assess political violence and its influence on governance. Highest scores go to countries with no armed opposition, and where the government does not indulge in arbitrary violence, direct or indirect. Lowest ratings go to civil war torn countries. Intermediate ratings are awarded on the basis of the threats to the government and business: whether the acts of violence have a political objective or not, whether violent groups represent a sizeable minority or not, how well organized these groups are and how much popular support they receive, how frecuent the act of violence are. and whether thev are neonraphicallv limited or not. Government Effectiveness Govemment Stability. Measures the government's ability to carry out its declared programs, and its ability to stay in office. This will depend on issues such as: the type of governance, the cohesion of the government and governing party or parties, the closeness of the next election, the government's command of the legislature, and DoDular aooroval of the oovernment Dolicies. Bureaucratic Quality. Measures institutional strength and quality of the civil service, assess how much strength and expertise bureaucrats have and how able they are to manage political alternations without drastic interruptions in government services, or policy changes. Good performers have somewhat autonomous bureaucracies, free from political pressures, and an established mechanism for recruitment and traininq. 50 Table A13, Cont'd Political Risk Services: by Agarecate Governance Indicator. Cont'd Regulatory Quality Investment Profile. Includes the risk to operations (scored from 0 to 4, increasing in risk); taxation (scored from 0 to 3). repatriation (scored from 0 to 3); repatriation (scored from 0 to 3) and labor costs (scored from 0 to 2). They all look at the qovernment's attitude towards investment. Rule of Law Law and Order. The Law sub-component is an assessment of the strength and impartiality of the legal system, while the Order sub-component is an assessment of popular observance of the law (thev are assessed Control of Corruption Corruption. Measures corruption within the political system, which distorts the economic and financial environment, reduces the efficiency of government and business by enabling people to assume positions of power through patronage rather than ability, and introduces an inherently instability in the political system. The most common form of corruption met directly by business is financial corruption in the form of demands for bribes connected with import and export licenses, exchange controls, tax assessments, police protection or loans. This measure is also concerned with actual or potential corruption in the form of patronage, nepotism, job reservation, "favor-for-favor", secret party funding, and suspiciously close ties between politics and business. The major risk arising from corruption is that a major political scandal produces a popular backlash resulting in a fall or overthrow of the qovemment, a maior reorqanizina or restructurinq of the country's political institution, or Table A14 The World Business Environment Survey (WBS) http://wwwl.worldbank.orq/beexUpart ii csai/csai maina_wbes.htm The World Business Environment Survey (WBS) is a survey conducted by the World Bank in collaboration with several other institutions. It is designed to provide information on the business environment facing private enterprises. It was conducted during 1999 and 2000 in 81 countries. The respondents were managers of firms in at least 100 firms per country. This survey asks several questions similar to those in the 1997 World Development Report survey that we used in constructing the 1997-98 version of the indicators. We therefore treat the WBS as the continuation of this source. The component of the WBS covering transition economies is referred to as the Business Environment and Enterprise Performance Survey (BPS), and surveyed 3000 firms in 20 transition economies. The questionnaire for this region included more detailed questions about corruption issues, including questions an "state capture" referring to the manipulation of the institutions of the state for private gain on a grand scale. Since these auestions were not included elsewhere, we use the BPS state capture questions as a separate source. In the table below we list the variables included in each of the govemance indicators. The World Business Environment Survey: by Aaqnreaate Govemance Indicator Voice and Accountability Business have voice to express Business are informed Political Stability Political instability Government Effectiveness Quality of customs Quality of public works Quality of power company Quality of Water Quality of public health Quality of public education Quality of central government Quality of central bank Efficiency of qvt in deliverinq services Regulatory Quality Regulations on starting new businesses Price controls Regulations on foreign trade Foreign currency regulations Goeneral uncertainty about requlations Rule of Law Corruption of bankers Quality of the Police Orqanized crime Street crime Courts-- fair & impartial Courts-affordable Courts-consistent Court's enforceability Confidence in judicial system today in insuring property rights General constraint-functioninq of the iudiciarv Obstacles to competition-violation of patents Quality of courts Control of Corruption Frequency of additional payments Dishonest courts Corruption as obstacle to business Bribery (% of Gross revenues) State Capture (BPS) Table A1 5 Institute for Management Development (WCY) http:/Iwww.imd.ch. The Institute for Management Development is an research and educational organization based in Lausanne, Switzerland. It has published the World Competitiveness Yearbook since 1987. Until 1996, this was a joint effort with the World Economic Forum. The World Competitiveness Yearbook analyzes the competitive environment in 47 countries. It is based on both objective data and surveys of perceptions. The survey questions over 4,000 local and foreign enterprises operating in the countries under analysis. Mean scores on the survey questions are reported in the yearbook for all countries. In the table below we list the questions In the table below we list the variables included in each of the governance indicators. We use data from the 2001 edition of the World Competitiveness Yearbook Institute for Management Development by Aggregate Governance Indicator Voice and Accountability Transparency Political Stability Risk of political instability Government Effectiveness Effective Implementation of government Decisions Bureaucracy as obstacle to business development Political Interference The government adapts its policies to changes in the economic environment Distribution of infrastructure goods and services is generally efficient. Regulatory Quality Legal Regulation of Financial Institutions Protectionism as obstacle to imports Control on foreign ownership Obstacles to foreign bidders on public contracts Political system as obstacle to development Real personal taxes as burden to work initiative Real corporate taxes as burden to entrepreneurship Legal framework as obstacle to competitiveness Customs as burden for international trade Price controls Competition laws as obstacles to competition Rule of Law Parallel economy as obstacle to business development Extent of tax evasion Confidence in fair administration of justice Confidence that person and property are protected Control of Corruption Improper practices in the public sphere 53 Policy Research Working Paper Series Contact Title Author Date for paper WPS2754 Revealed Preference and Abigail Barr January 2002 T. Packard Self-insurance: Can We Learn from Truman Packard 89078 the Self-Employed in Chile? WPS2755 A Framework for Regulating Joselito Gallardo January 2002 T. Ishibe Microfinance Institutions: The 38968 Experience in Ghana and the Philippines WPS2756 Incomeplete Enforcement of Pollution Hua Wang January 2002 H. Wang Regulation: Bargaining Power of Nlandu Mamingi 33255 Chinese Factories Benoit Laplante Susmita Dasgupta WPS2757 Strengthening the Global Trade Bernard Hoekman January 2002 P. Flewitt Architecture for Development 32724 WPS2758 Inequality, the Price of Nontradables, Hong-Ghi Min January 2002 E. Hernandez and the Real Exchange Rate: Theory 33721 and Cross-Country Evidence WPS2759 Product Quality, Productive Aart Kraay January 2002 R. Bonfield Efficiency, and International Isidro Soloaga 31248 Technology Diffusion: Evidence from James Tybout Plant-Level Panel Data WPS2760 Bank Lending to Small Businesses George R. G. Clarke January 2002 P. Sintim-Aboagye in Latin America: Does Bank Origin Robert Cull 37644 Matter? Maria Soledad Martinez Peria Susana M. Sanchez WPS2761 Precautionary Saving from Different Richard H. Adams Jr. January 2002 N. Obias Sources of Income: Evidence from 31986 Rural Pakistan WPS2762 The (Positive) Effect of Norbert R. Schady January 2002 T. Gomez Macroeconomic Crises on the 32127 Schooling and Employment Decisions Of Children in a Middle-income Country WPS2763 Capacity Building in Economics: Boris Pleskovic January 2002 B. Pleskovic Education and Research in Transition Anders Aslund 31062 Economies William Bader Robert Campbell WPS2764 What Determines the Quality Roumeen Islam January 2002 R. Islam of Institutions? Claudio E. Montenegro 32628 Policy Research Working Paper Series Contact Title Author Date for paper WPS2765 Inequality Aversion, Health Adam Wagstaff January 2002 H. Sladovich Inequalities, and Health Achievement 37698 WPS2766 Autonomy, Participation, and Learning Gunnar S. Eskeland January 2002 H. Sladovich in Argentine Schools: Findings and Deon Filmer 37698 Their Implications for Decentralization X WPS2767 Child Labor: The Role of Income Rajeev H. Dehejia January 2002 A. Bonfield Variability and Access to Credit in a Roberta Gatti 31248 Cross-Section of Countries WPS2768 Trade, Foreign Exchange, and Energy Jesper Jensen January 2002 P. Flewitt Policies in the Islamic Republic of David Tarr 32724 Iran: Reform Agenda, Economic Implications, and Impact on the Poor WPS2769 Immunization in Developing Countries: Varun Gauri January 2002 H. Sladovich Its Political and Organizational Peyvand Khaleghian 37698 Determinants WPS2770 Downsizing and Productivity Gains Martin Rama January 2002 H. Sladovich In the Public and Private Sectors Constance Newman 37698 of Colombia WPS2771 Exchange Rate Appreciations, Labor Norbert M. Fiess February 2002 R. Izquierdo Market Rigidities, and Informality Marco Fugazza 84161 William Maloney