Jon Chait has a great piece on Alan Reynolds’ “research” into inequality, noting that the main point here isn’t to convince anyone that Reynolds is right (the work would need to be less amateurish than that), but simply to convince ordinary people that there’s a big, complicated, confusing controversy among the experts on this subject so who’s to say:
For example, [Reynolds] argues that Piketty and Saez’s data does not account for the massive rise of tax-sheltered pensions, such as 401(k) plans, which are “invisible in tax return data.” Because 401(k) plans are now common among middle-class earners, tax returns miss a huge source of their wealth and thus make them look misleadingly poor. This sounds sensible enough, but it is wrong on several levels. 401(k)s didn’t just appear out of nowhere; they mostly replaced defined benefit pensions. And, like the old pensions, 401(k)s do appear on tax returns when the accounts are withdrawn. On top of that, economists think most taxfavored assets are concentrated in the hands of the rich anyway, so, even if Reynolds were right about tax returns, it would very likely make inequality look even worse.
But whether the missing data would make inequality look worse or better is really beside the point. Reynolds’s role is merely to point out that the data is imperfect. The skeptic challenging the expert consensus must be fluent enough in the language of the experts to nibble away at their data. (The evolution skeptic can find holes in the fossil record; the global-warming skeptic can find periods of global cooling.) But he need not–indeed, he must not–be fluent enough to assimilate all the data himself into a coherent alternative explanation. His point is that the truth is unknowable.
Reynolds also turns out to be what you might call a fake economist. Obviously, there’s a somewhat difficult asymmetry here. If you’re an egalitarian liberal, you’ll want to see the government implement anti-inequality policy insofar as you think a troubling level of inequality actually exists. So I would not, for example, urge Iceland to adopt new anti-inequality policies because Iceland’s income distribution already is highly egalitarian. In the American context, however, new pro-equality policies are needed. Deciding what you want to do, in short, requires you to actually know something about the structure of wealth and income distribution. This, in turn, requires a certain level of analytic caution and modesty that makes it hard to write bombastic screeds on the subject.
Reynolds and his ideological fellow-travelers, by contrast, don’t think we should try to reduce inequality and their commitment to that position is completely independent of their assessment of the empirical facts regarding the extent of inequality. So if Reynolds is getting all his data wrong, nobody on his side is really going to care about that, since, from the right-wing point of view, the question Reynolds is asking doesn’t actually matter except as a tool in public debate.