I think this is a useful corrective to an often misleading stereotype, but I also think that both the original stereotype and the corrective are stuck in the wrong frame of thinking - foregrounding class and income when the big drivers of populism are education, age, identity
In other words, while there are indeed plenty of "provincial middle class populists", their relative wealth and middle class jobs aren't really predictors of their populism at all. Its their nationalism, their social conservatism, their age, they educational experience
These are the salient features that emerge for analysis after analysis as setting apart populists (or more accurately radical right supporters) from non-populists, be they "left behind" factory workers or middle class provincials.
There's also a more subtle problem at work here - the conflation of relative and absolute rates of support. This is very evident in much of Dorling's claims in this area. Even if middle class provincials aren't partcicularly prone to back rad rt, there will be lots of such
voters among the rad rt support base simply because *most people* are middle class or provincial (or both). Whereas even if poorer factory workers are much more likely than others to back rad rt, most rad rt voters won't come from that group because its too small.
An analogy - African Americans are much, much more likely to back Dems in US, and whites break modestly away from Dems , but most Dem voters are white because most *voters* are white. Blacks are *relatively* v over-represented, but whites are *absolutely* the much larger group.
