A story in yesterday’s Washington Post about unlimited individual campaign donations included this interactive bubble chart. In it, donation data for Democratic and Republican donors are encoded into bubbles and grouped together. A comment below the article asked the following:
So, is the bubble chart the best way to show the data? Does the interactivity help? How can it be improved?
I pulled the data from this Google spreadsheet provided by OpenSecrets.org in a write-up posted yesterday.
The spreadsheet data seems a little screwy. Some donors, such as Schwab gave more to individual parties than their “Total” giving, and some gave less. The latter might be for donations to other parties, but I can’t figure out the former.
The WP bar chart over time is a nice aggregate view. One of the goals of the bubble chart must be to show individuals along with the aggregate. I don’t see any purpose to the interactivity except entertainment. Rectangles aggregate better than circles, so I think a treemap would be better. Individual sizes and counts are still not easy to compare (as they would be for a bar chart), but I assume compactness is another goal.
I used the excess amount for size instead of the total amount, since that’s what should be aggregated and matches the bar chart. That is, the point of the story is how much each party has benefited from the excess giving. The aggregate ratio changes a little. Instead of $33.3M to $15.6M, for the excess it’s $8.5M to $3.2M.