Last month, while World Bank President Jim Yong Kim launched the gender data portal, U.S. Secretary of State Hillary Clinton remarked that “data not only measures progress, it inspires it”. Indeed when data is both relevant and effectively communicated, it can help to inform policies, identify challenges, and catalyze changes and innovations that deliver development results.
With that goal in mind, we started an Open Data Lab. One of our objectives is to help the development community become more effective data communicators by experimenting with different data visualization techniques and tools. The human brain finds it easier to process data and information if it is presented as an image rather than raw numbers or words. And visualizations that let and encourage users to interact with data can deepen their understanding of the information presented.
We decided to begin with data that defines our challenging work and mission: poverty reduction. Over 20 years ago Martin Ravallion and Shaohua Chen with various collaborators along the way developed the now widely used methods, tools and indicators to monitor the population “living on less than $ a day”. Earlier this year they released the latest update of the World Bank’s global poverty monitoring data along with a re-vamped version of PovcalNet: an interactive on-line computational poverty analysis tool. We downloaded the table with regional poverty estimates and used PovcalNet to calculate the results for additional poverty lines. We then combined all the data in the three-panel visualization embedded below.
Click here for a guide on using this data visualization.
You probably already know that the global incidence of extreme poverty (defined as living on less than $1.25 a day) has fallen from 52 percent in 1981 to 22 percent in 2008. And three regions in particular – East Asia and the Pacific, South Asia and Sub-Saharan Africa – have experienced very different poverty trends over this period. But did you know that:
• Between 1981 and 2008, three regions (East Asia and Pacific, South Asia and Sub-Saharan Africa) have together consistently accounted for over 95 percent of the developing world’s population living on less than $1.25 a day?
• Even if we doubled the extreme poverty line to $2.5 a day, these three regions would still account for 93 percent the poor in the developing world?
• Even if we doubled the poverty line further to $5 a day, 88 percent of the developing world’s poor population would be still found in these regions.
You can explore more interactive country and regional data dashboards with the latest World Bank poverty statistics here: http://povertydata.worldbank.org/poverty/home
What do you think? Did you learn or discover something new about global poverty data? Did it help improve your understanding of these data?
It was excellent method to present data but I was surprised when gone through the data of South Asia, it has shown increasing number of poverty but states of this regions are claiming decreasing number of poverty such as in Nepal lastest data shows only 24% people are under poverty.
We do not curate indicator data on such housing characteristics for different countries. But do maintain a library with household survey microdata and DDI compliant metadata which you can search at the variable level: http://microdata.worldbank.org/index.php/home. This will likely turn up some data that you will find useful.
Great no doubt!
Three questions though:
1. I know there is a long methodology somewhere but to help a lay person understand, are the numbers derived from only those reported or were the surveys exhaustive? In other words, how representative of the total poupulations per continent are these numbers and what were the margin of errors? Sampling has its limitations.
2. What has been the actual demonstrable impacts of releasing this kind of data on poverty reduction itself for the Bank and the countries themselves? Do we know?
3. Do you have similar analysis for poverty per capita? Aggregate numbers generally have the tendency to mask 'real' progress (consider the difference in differences) or does the Bank even care?
It looks like Africa gets to get the flack for poverty yet only accounts for only 30% of global poverty. Why doesn't the Bank make this absolutely clear to the global development community?
Keep us posted on even more progress with this great tool!
You are welcome Lori. I like the video and pleased this will help others. Watch this blog... more coming soon!
It can be a very helpful tool. Great work.
A new day of collaberating is unfolding and we should all benefit as we contribute to a brighter more intelligent discussion with new portable powerfull tools.
Wow! great data analysis and visualization!
Great work, very useful now that poverty targetting is the most important order of business. Can you try something similar with inequality? Ginis to begin with. And with vulnerability? This is trickier to define, but to keep it simple, can we start with the pop, say 20% above pov line? Now that food price increases are the second more importnat order of business, this could come handy.
Nice! Great job!! Its export function is quite useful. I exported a view in PDF and open it with Adobe Illustrator for creating a new infographic image by jusxtaposing with photos, images and figures. Then I placed it in a PPT slide. Lots of ways to use in various communication products! Like it!
Great work Johan and Juan. Would love to see what this inspires from others also. How are you sharing these visuals to make sure they inform and inspire other developers and visualizers?
Love what you're doing with Open Data and the power of pictures.
Hey Johan, great job. In case you didn't see it, this was selected as Tableu's Viz of the Day - very well done! http://www.tableausoftware.com/public/community/viz-of-the-day
Thanks Amparo. Yes we'll add inequality on our to-do-list. Using the Mean Log Deviation (MLD) measure, as outlined in Shaohua and Martin’s recent working paper “more relatively-poor people in a less absolutely-poor world”, is a good approach for global and regional aggregation/decomposition of inequality. And note that PovcalNet allows you to quickly get one measure of vulnerability by varying the poverty line by x%. A $1.25 per day is equivalent to $38 per month ($1.25*365/12) and 20% increase would mean setting the poverty line equal to 45.6. Enter this poverty line in PovcalNet here and let us know what you find.
Congratulations Johan, Juan & Team! A fantastic way to provide greater insight into data what would otherwise be lost in an excel sheet...
If a may, I'd like to plug in the work of our team at the Independent Evaluation Group (IEG), using Tableau to make our data on project performance assessements more accessible, at: http://ieg.worldbankgroup.org/content/ieg/en/home/ratings.html
I agree with all those above who praise you for releasing this visualization tool. And great quote from Hilary Clinton.
Given that your analysis focuses on the impact of the different poverty lines, it would be nice to show it in one graph, rather than the five separate graphs that are not visible at once. Also, in your inequality graph on the "Poverty & Equity Data" home page, I would suggest that the scale should not change throughout the years so that we can rely on the length of the bars to track overall levels of inequality through the years.
Thanks for putting these tools out there.
Francis: Thank you for the excellent suggestion. We'll fix that!
Thanks Alex. Your dashboard looks interesting. We'll try it out!
Tremendous to see how Tableau now lets anyone quickly create dynamic charts that approach Gapminder's famous visualizations for the UNDP's Human Development Report back in 2005 (http://hdr.undp.org/external/gapminder/2005/2005.html). Very cool.
I was also really impressed by your redesigned DataBank interface - the World Bank is doing a great job in general of making your data visually accessible and engaging. Would it be possible to hear from the graphic designers about the challenges they faced in updating the visualizations - solutions they were pleased with, limitations they had to live with, etc. - either through a blog post here, or by sharing who they were so we could follow up with them?
Thanks Aleem. We'll do that! And happy to hear you think we're off to a good start. Got any ideas going fwd? Let's brainstorm over lunch one of these days.
Thanks Nathan. Good to hear you like our spruced up DataBank interface and data sharing functionalities. We'll follow up with graphic design folks. And glad to see you are keeping busy in Namibia. Your recent map of Namibian childcare facilities mashed up with OVC data is very informative.
Great job! I own perception reflects your research, I find it very convienent to understand the point.
Dear Johan Mistiaen,
I would like to know if the World Bank collects data about the average useful floor area per dwelling and per person, and whether it can be published in future. I'd really appreciate it
Good point, and yes, the poverty statistics shown in the graphs have taken into account the inflation. The $1.25-a-day poverty line is at the 2005 price level. Consumption/income data used to calculate poverty rates has been adjusted to the 2005 price level, and it has also been adjusted for different purchasing power of local currencies across country. For more information on the methodology of poverty estimation, please refer to PovcalNet - an online poverty analysis tool developed by the World Bank, (http://iresearch.worldbank.org/PovcalNet/index).
The global absolute poverty rate measured at $1.25 a day has fallen from 52.2% in 1981 to 22.4% in 2008, despite the population growth.
The first chart shows the share of the poor population in each region in the total poor population in the developing world. The number for South Asia is rising because population growth outpaces poverty reduction. There is no doubt the poverty rate in South Asia has declined.
Great job, but does this graph take into account the inflation over the period? $1.25 in 1981 would have the same purchasing power as more than $2.50 in 2008. If this inflation has not been accounted for then 3 billion people are living in the same poverty now that 2 billion people were living in back in 1981.
The population of the World was about 4.5 billion in 1981 compared to 7 billion now, so it would appear that there has been little or no improvement over 30 years with more than 40-45% living in absolute poverty
Johan, thanks for letting us use your visualization to show how global filters and highlighting work. Here's the link. http://www.youtube.com/watch?v=FEqESQ2LCXU Look forward to seeing what you make next. Cheers!
thank you for this information I love to read this post because it is very useful and easy to understand, after reading this page I will see see the other pages of this web hopefully have a lot of good information
Blog posting is very common now a days. As readers want quality posts so certain things should keep in mind. You shared very helpful tips. I will recommend this site to my other friends also. Thanks for sharing.
This could be one particular of the most beneficial blogs We’ve ever arrive across on this subject. Basically Magnificent. I am also a specialist in this topic therefore I can understand your hard work. Jantung Bengkak
poverty, and now inequality, is a very interesting though complex subject. I am trying to have a regional approach to poverty reduction, and we are caught up in the issues of measurements used by different Member States in SADC.