Housing Finance in Nigeria
This is a report from the World Bank Group on housing finance in Nigeria. From the introduction:
"This report summarizes the results of the analytical work on housing market finance carried out by the World Bank Group at the request of the Ministry of Finance. The purpose of the work was to inform the policy dialogue about how best to develop a sustainable housing finance market in Nigeria, and improve the effectiveness of interventions aimed at stimulating the housing market and providing quality housing to the population."
It contains a lot of data and presented several opportunities for a wide variety of data visualizations.
The radar chart is often maligned and rarely used properly. One of its issues is that its corners do not relate to an intuitive or performance ranking of the categories. Also, the comparisons are difficult across categories because of the lack of alignment.
The present data offered a rare opportunity: it is organized along the six cardinal points. It might be easier to understand it when organized as such, than in a table or in a bar chart. After experimenting with different formats, we settled on a version that shows the national average in a pale dotted line, superimposing the regional data in dark blue. From our tests, the connecting lines appear to make it easier to see at a glance which regions are above and which are below the average.
The opportunities to use a parallel coordinates graph are rare, but here the data was fitting. For more details on how to make one in Excel, see the blog post.
Other examples of slopegraphs with similar data.
Slopegraphs are an effective way to show a before and after measure, especially when there is not enough data for a detailed time series that would show the complete evolution.In this case, the poverty rates are compared between two time periods, and across two measurements.
In the next example, the data was originally presented as clustered bar charts, in no particular order. Using a slopegraph immediately put the data in order of performance and allowed to compare the rate of increase, especially how one series is catching up with another.
Dot plots can be useful when there is little overlap across categories. In the present case, only the tall blue bars were easy to compare across categories. With the dot plots, all four series become clear. The "Health" series changes location, to the left, rather than becoming so small it is almost invisible in the original bar charts. Notice how the main topic, house-related expenditures, is made more visible, to serve as a visual reference.
In the following example, it becomes clear that the shape of the household average is different from that of the per capita average. It was nearly impossible to notice from the original table.
The original graph combined two quantitative series for each income group: the proportion of ownership and the share of the population. The original graph was an attempt at a Marimekko chart, but without the right proportions. The triangle, notably, did not convey any data and it clashed with the row height.
The concept was preserved, but the execution was fixed.
This is a simple fix. The original graph did not differentiate the actual data from the projections.
The original flow chart was confusing. It was not clear to the eye that the central piece, the one explained by the chart, was the liquidity facility. The use of colours is also confusing and inconsistent, forcing the reader to decode it each time. There are also certain items, like the dark blue ovals, that serve no purpose and others like the dashed line that are simply not necessary. It takes some time to even realize this is a flow chart that follows money between different entities.
In the new version, the liquidity facility is made the most visible. The two flows (left-to-right, right-to-left) are shown using a consistent position and colour. The circles represent the entities and their size hints at their real-life sizes, where borrowers are smaller and more numerous.