I appreciate the effort that went into making this, but I get really annoyed with analyses like this that show heatmaps comparing highly heterogeneous groups (i

I appreciate the effort that went into making this, but I get really annoyed with analyses like this that show heatmaps comparing highly heterogeneous groups (in this case, states, income groups, age groups, and education groups) that use an absolute scale. Of course absolute migration looks extreme in NY and Texas compared to, say Missouri or Iowa. This would be a much more interesting story if they used a lift metric that controlled for state/group size and/or state/group migration volume.