National Geographic features this graphic illustrating migration into the U.S. from the 1850s to the present.
What to Like
It's definitely eye-catching, and some readers will be enticed to spend time figuring out how to read this chart.
The inset reveals that the chart is made up of little colored strips that mix together. This produces a pleasing effect of gradual color gradation.
The white rings that separate decades are crucial. Without those rings, the chart becomes one long run-on sentence.
Once the reader invests time in learning how to read the chart, the reader will grasp the big picture. One learns, for example, that migrants from the most recent decades have come primarily from Latin America (orange) or Asia (pink). Migrants from Europe (green) and Canada (blue) came in waves but have been muted in the last few decades.
Initially, the chart is disorienting. It's not obvious whether the compass directions mean anything. We can immediately understand that the further out we go, the larger numbers of migrants. But what about which direction?
The key appears in the legend - which should be moved from bottom right to top left as it's so important. Apparently, continent/country of origin is coded in the directions.
This region-to-color coding seems to be rough-edged by design. The color mixing discussed above provides a nice artistic effect. Here, the reader finds out that mixing is primarily between two neighboring colors, thus two regions placed side by side on the chart. Thus, because Europe (green) and Asia (pink) are on opposite sides of the rings, those two colors do not mix.
Another notable feature of the chart is the lack of any data other than the decade labels. We won't learn how many migrants arrived in any decade, or the extent of migration as it impacts population size.
A couple of other comments on the circular design.
The circles expand in size for sure as time moves from inside out. Thus, this design only works well for "monotonic" data, that is to say, migration always increases as time passes.
The appearance of the chart is only mildly affected by the underlying data. Swapping the regions of origin changes the appearance of this design drastically.