James’s list is cherry picked. If you’re going to do it by HDI, then this is the list:

Norway Switzerland Australia Germany Denmark Singapore Netherlands Ireland Iceland Canada United States New Zealand Sweden United Kingdom Japan South Korea Israel Luxembourg

He included Austria, Belgium, Finland, France, Greece, Italy, Portugal, and Spain, which aren’t HDI top 20, and excluded Israel, Luxembourg, Singapore, Switzerland.

I don’t agree with his intentional exclusions, because if guns/cap and gun homicide were related, they’d be related whether the country was big or small.

Weighting by population is a wretched way to look at it, because it puts all the weight on the USA, which is an outlier. If anything, we should be weighting against population density, which makes his exclusion of Singapore especially curious.

If we wanted some way to normalize things by population, we’d have to break the US into states, and include states in the plot with HDI countries. When we did that, we’d get basically 20 more data points to throw into the US states graph, which already has no correlation.

Gun homicide within US states is primarily correlated with GINI coefficient and percent black population, on a multivariate analysis. I highly, highly suspect that most of the HDI countries in the comparison reflect much different wealth equality and black population than we do here, and if we were to run the same sort of multivariate analyses that account for those here, we’d see those account for those there as well.