Showing posts with label graphs. Show all posts
Showing posts with label graphs. Show all posts

11 June 2025

Chart of the Day: What do Africans think their governments should be doing?

Afrobarometer asked over 33,000 Africans between 2010 and 2012 what the most important problem facing their country that government should address is. Here are their answers. With apologies for the tiny font, but it's worth reading down the full list (I left off a few of the country-specific responses at the bottom).

Data from: Benin 2012, Botswana 2012, Burkina Faso 2012, Burundi 2012, Cape Verde 2011, Ghana 2012, Kenya 2011, Lesotho 2012, Liberia 2012, Malawi 2012, Mali 2012, Mauritius 2012, Namibia 2012, Nigeria 2012, Sierra Leone 2012, South-Africa 2011, Tanzania 2012, Togo 2012, Uganda 2012, Zimbabwe 2012 (Base=33598; Weighted results)


I'm quite surprised by how high up water supply is, but less surprised by the top 3 of unemployment, the economy, and poverty. The public policy challenge is still, first and foremost, about broad-based inclusive economic growth. Interesting to compare this with Justin Sandefur's analysis of what African researchers care about (jobs).

The tragedy is that we don't really have a clue what policy instruments can create jobs. For most of sub-Saharan Africa the challenge is a lack of demand for labour. What is needed is a way of linking African workers with consumers who have money - who are mostly in rich countries. This link could come in 3 ways:

1: Trade. Africans stay where they are and export things to rich countries. This one looks difficult in most countries, which are uncompetitive with poor Asian countries in manufacturing, and don't yet have the skills or infrastructure for high-tech service exports. Gains to agricultural productivity holds some promise, but faces serious barriers to getting going.
2: Migration. The Africans come to rich countries. An economic no-brainer, and a political non-starter.
3: Tourism. The rich people go to Africa. Tourism? Really?

There will probably be marginal improvements in all 3 areas, but its hard to see where the really big shift that could get millions of Africans up to rich country poverty lines of around $12.50 per day over the next generation is going to come from.

The very easy to use online Afrobarometer data analysis tool is here.

[and before anyone says it, of course Africa is not a country, but actually the patterns look pretty similar when you look at the country-level data, I just couldn't figure out a good way of showing that data visually - very open to suggestion]

03 April 2025

Bad Graphics

This is a guest post by Sean Fox at the LSE

This infographic, which came to my attention a few weeks ago on International Women’s day, has been on my mind because it is one of the WORST visual presentations of data I have seen in years: 



So what? Well, it contains information on an interesting and important topic (attitudes about domestic abuse) in a UN report. It should inform. Instead it confuses and distorts the facts. It violates almost every rule outlined in the bible of infographics, The Visual Display of Quantitative Information by Edward R. Tufte. Let me just name a few.
  1. It looks like a quasi-pie chart. As such it implicitly suggests to the viewer that the slices represent portions of a whole. They do no such thing. They represent survey responses from a relatively small and arbitrary selection of countries around the world. 
  2. The sizes of the ‘slices’ do not correspond to the numbers they purportedly represent. Just compare the Rwanda slice to the Vietnam slice. Huh?? 
  3. It uses multiple colours. This is a great way to pack more data into a small space, but in this case the colours actually contain no information at all. They’re just randomly assigned. More visual confusion.
  4. It uses a lot of ink to represent a small amount of data. Rule number 1 of good info graphics is to maximise the data/ink ratio. Less is more. 
So, how should it have been presented? There are many better ways, but a very simple one, which took me about 5 minutes in Excel is this:



While the first figure confuses the brain and obscures the significance of the data, this simplified version immediately throws up all kinds of interesting questions. Why do the women of the post-Soviet nations of Serbia, Georgia and Kazakhstan seem to have some of the lowest tolerance for domestic abuse in the world? How is it that the women of Jordan, which has a relatively liberal and modernising king and a female role model in the politically active and globetrotting Queen Rania, seem to largely accept domestic violence? What accounts for the wide gap in attitudes between women in the East African nations in Ethiopia and Rwanda? Is it due to “culture” or government policy and discourse?

These are interesting and important questions that are revealed by a simple improvement in the presentation of the data.

Come on, UNICEF. You can do better.

17 February 2025

Chart of the day: Evidence-based aid in the UK

This chart from the LSE "Impact of Social Science" handbook shows a ranking of UK government departments by the number of references to academic research found on their websites. DFID comes third.

14 February 2025

A consultant's love life

An oldie but a goodie, if you haven't seen this yet, I think it is probably better than most of the other economist-chart valentine's jokes doing the rounds.

Romance: A BCG Analysis
HT: Steve

17 May 2025

What is better than deworming for increasing school student attendance?

For me this was the stand-out chart from Esther Duflo’s recent TED talk. Apparently providing information on the returns to education does even better than deworming for increasing school attendance.

Student Attendance

I am somewhat surprised that there haven’t been more efforts to replicate the Kenya deworming finding, which is quite old now (noting that there is one study on India here - but what about other African countries?). A key defence of the randomistas position is that external validity can be improved by replication of findings in different settings. The incentives probably aren’t there for academics to do boring replication studies, but surely there is an incentive for donors to want to fund such studies?