The graph is almost an essay in "how little raw data can tell you". Are the areas of more women from emigration, age, areas from which many men have been taken to prison, something else?