Sailesh Tiwari
1818 H Street, NW
Washington DC 20433

Maura Leary
1818 H Street, NW
Washington DC 20433

Resources: FAQ



What is “Visualize Inequality”?

Visualize Inequality is an interactive repository of all data on inequality of opportunity produced by the World Bank.


How many countries are covered in Visualize Inequality?

In its most current version, the Visualize Inequality dashboard covers 68 countries in Sub-Saharan Africa, East Asia, Europe and Central Asia and the Latin America and Caribbean regions. The only regions that are not currently covered are South Asia and Middle East and North Africa, but work is ongoing and the data for the region will be included once it is complete. We also use the data on the most recent PISA tests (2009 and 2012), to provide data on educational achievement for 68 countries and economies.


What kind of inequality data can I expect to find here?

Data on inequality of opportunity, which allows users to go beyond country level averages on access to basic opportunities– such as quality education, nutrition, clean drinking water, adequate sanitation, electricity, etc.– and observe a measure of the differences in access to these opportunities among children in a particular country given their gender, parental and other socio-economic characteristics.


But what about data on conventional measures of inequality such as Gini coefficient?

We do not currently include data on inequality of outcomes such as income, labor status or household expenditures. The dashboard focuses on inequalities that are out of the control of individuals, and that over the course of life may translate into inequalities of outcomes. Therefore, we focus on children and explore differences in access to basic services that are important to enjoy full and healthy lives –such as access to education, health services, water, electricity and the like.


How do you define “inequality of opportunity”?

There is equality of opportunity when all children in a particular country have an equal chance at a better life through equal access to basic services such as quality education, health and infrastructure facilities irrespective of their circumstances. Viewed this way, equality of opportunity is about leveling the playing field for all. Thus, there is inequality of opportunity if access to these basic services is dependent on their circumstances.


How do you measure inequality of opportunity?

There are several ways of measuring inequality of opportunity, but here we focus exclusively on a methodology that is known as the Human Opportunity Index (HOI). The HOI shows the extent to which individuals of different circumstances have varying access to opportunities such as school attendance, immunization, access to water, sanitation etc. A summary measure of these differences is the dissimilarity index, or the inequality index. This inequality index can be interpreted as the share of opportunities that need to be reallocated from one group to another for there to be equality. For example, consider the following situation:

  • Country A: 40 of the 50 children in urban areas are enrolled in school and only 20 of the 50 children in rural areas are enrolled.
  • Country B:  35 of the 50 urban children are enrolled in school and 25 out of the 50 rural children are enrolled.

Coverage rates

Country A
(100 children)

Country B
(100 children)

Urban (50)

40/50= 80%

35/50= 70%

Rural (50)

20/50= 40%

25/50= 50%


60/100= 60%

60/100= 60%

When you look only at the coverage rate – that is the percentage of children that go to school – the number looks identical for the two countries.


Country A
(100 children)

(100 children)

Urban (50)



Rural (50)






But when you consider the dissimilarities, you will note that Country A is much more unequal than Country B. Please see the documentation files for a more detailed explanation.


What do you mean by circumstances?

The word “circumstance” has a very specific meaning in this context and refers to characteristics that a person is born with and has no control over. For example, gender, place of birth, parental socio-economic status etc. are all examples of circumstances. These are also characteristics that should not hamper a child’s access to basic opportunities.


Can you ever observe all of the circumstances in any data set? Doesn’t that affect your calculations?

Ideally, in order to calculate the true measure of inequality of opportunity, we would use all possible circumstances that may affect the access to basic opportunities for children. However, not only is it hard to know what all the relevant circumstances for any opportunity and society are, but it is also impossible to find datasets that capture all of them. Therefore, in each case, we work with the available information. However, the inequality index mitigates this concern to some extent. If any additional circumstance is added to the calculation, the measure of inequality can only go higher. This implies that the measure that we currently use is the lower-bound of the true inequality of opportunity.


Once you calculate inequality of opportunity, can you break down the share that is attributable to each of the circumstances?

Yes, a technique called the Shapley Decomposition (see documentation and resources for more details) can be used to calculate the relative contribution of each circumstance to the inequality index. In other words, it is possible to pin down the extent to which gender or the area of residence (urban/rural), for instance, contribute to the total disparity in access to water.


What is the Human Opportunity Index?

The Human Opportunity Index (HOI) is the inequality-adjusted measure of access to basic services. Take the example of primary school enrollment. Let’s say that two countries have exactly the same overall enrollment rates, but in one country, the majority of those enrolled are wealthy boys in urban areas, whereas in the other, enrollment is distributed nearly equally among different socio-economic groups. The country with the more unequal access (majority wealthy urban boys) would be “penalized” at a higher rate in comparison to the other, so the HOI for primary school enrollment in that country would be lower than the HOI for the country with more equal distribution—even though the two countries have equal overall enrollment rates.


How is the Human Opportunity Index different from conventional measures of access?

Conventional measures of access give you a raw estimate of the fraction of children who have a particular opportunity. Expressions such as “45% of the children in so and so country were fully immunized” or “86% of children had access to electricity” are all too familiar. The Human Opportunity Index adds an additional lens to look at those same numbers. Yes, 45% of children may have been immunized, but maybe most of them were systematically of a certain, privileged type; yes, 86% may have electricity, but what if they are disproportionately located well developed urban areas? This means that the raw estimates do not capture how these opportunities distribute amongst the population; it does not capture inequality in the access to opportunities. The HOI methodology is a rigorous way of bringing these “equality concerns” into the conventional indicators and understanding what is behind the national-level numbers.


What are the advantages of the Human Opportunity Index over other conventional measures of access?

The Human Opportunity Index goes beyond conventional measures of access to include inequality sensitivity. If a child’s circumstances do not affect his or her access to an opportunity, the HOI and coverage rate of an opportunity will be the same. However, if there are disparities then the HOI will be lower than the measure of access, with the difference between the two representing the “penalty” for unequal distribution.


How can you interpret changes in Human Opportunity Index over time?

Changes in Human Opportunity Index over time can be decomposed nicely into three components: the “scale effect”, the “equalization effect” and the “composition effect”. Any improvement that is uniform and universal for all groups and a result of policy interventions that “lift all boats” is registered as scale effect. Improvements that are disproportionately higher for the previously underserved groups through policies that benefit these groups (quotas and reservations for certain ethnic groups, building schools in remote areas and scholarships for girls etc.) are likely to be recorded as equalization effect. Changes in HOI that are due to fundamental change in the underlying characteristics of the population between the survey rounds is recorded as “composition effect”.


What datasets do you use?

We use several datasets. Demographic and Health Surveys (DHS) are the primary source of information for the Africa, South and East Asia regions, whereas harmonized regional household survey datasets are used for Europe/Central Asia and Latin America. We use PISA 2009 and 2012 to assess inequality of opportunity on educational achievement.


How come I can’t find data on my country here?

Visualize Inequality has been built in an extremely decentralized, bottom-up approach where we have sourced information from several regional exercises or reports. While data for the Latin America and Caribbean and Europe and Central Asia regions should be more or less complete in their coverage as they are based on regionally harmonized data, the information is less complete for other regions such as Africa where the information is based on a regional report which focused the analysis only on 20 or so countries for which recent DHS data was available.


How come I can’t compare my country in one region to a country in another region?

The definition of opportunities as well as the specific set of circumstances used to calculate inequality of opportunities is region-specific, thus in the current stage of this work, one can only make comparisons with countries in the same region. The reason for doing that is to ensure that the list of opportunities and circumstances used are tailored to the context and priorities of each region. For example, access to basic services that are relevant to most sub-Saharan African countries, for example, may be less so to a pre-dominantly middle-income region like Europe and Central Asia, and vice-versa.


I am a blogger, how do I share or use these visualizations in a blog?

Once you have created a particular visualization that you would like to use in your blog, you can embed the link directly into your blog.


I am a journalist, how do I use visualizations in my article?

Once you have created a particular visualization, the platform allows you to directly export that visualization as an image. You could alternatively also do a screen capture and extract the image using photo-editing software. Alternatively, if you are publishing exclusively on the web, you could directly embed the dashboard with the article.


I am a student/researcher, how do I access the underlying data used in the analysis here?

Several sources of data are used in what is presented here. The platform that is used allows the user to directly export the underlying data in an Excel format. Please see the tutorial video in the resources page for details.
If you are interested in unit level data, these are publicly available through sources such as DHS ( For cases in which we do not use such public access data, the process of accessing the raw data is a bit more involved. Please contact us if that is what you are looking for.


I am a student/researcher and would like to apply this methodology to a dataset from my country. What methodological guidance can you provide?

E-learning modules, PowerPoint files describing the methodology in varying degrees of details, all relevant STATA .ado files together with an illustrative application are included in the Resources page. If there are specific issues not directly addressed there, we would be happy to provide assistance. Please contact us.