Justice Map in development

You can see my early version of Justice Map.

This is for demonstrating what the layers look like, getting feedback on those, and not anything close to what it will look like when it is finished.

It is a set of open map layers based on the 2010 Census and 2011 American Community Survey for race and income. I put census tracts up online, but I plan to add counties (for both) and block groups and blocks (for race) in the near future.

Currently I've been exploring different ways of choosing categories. I'm pretty happy with the color schemes from colorbrewer2.org (though I've decided that I cannot use the blue or green ones since they are too close to what most map base layers look like - so I'm restricted to a smaller set of colors than I'd like).

But categorization is difficult as I want to balance having real differences between categories (and thus large enough percentage differences) and displaying the same amount of data in each category (as that makes the maps look nice and it maximizes the information it communicates).

You can see my extended thoughts on categorization at Stack Exchange

To make it even worse, the entire process is extremely biased by the fact that everything is effectively weighed by area. So a county with ten thousand people can take up a lot more map space than millions of people in a major city. One solution to this is to use a dot-based approach and the dot density indicates how many people are in an area (representing anywhere from 1 to 1000 people per dot depending on your zoom setting). I might eventually do that, but I'd like to see how someone else has done it as placing several million dots in several million polygons could take a lot of time.

The other way the process is biased is that I get different breaks if I analyze the data using census tracts, counties, block groups, or blocks. I chose to use census tracts.

It is a project of Sunlight Labs and Energy Justice.