Sunday photoblogging: yellow doors

by Chris Bertram on May 5, 2019

Metal panel

Evidence and conventional wisdom

by John Q on May 5, 2019

I’ve been looking over some posts from the bright dawn days of blogging in the early 2000s. One thing that struck me is that some ideas I put forward as unconventional but evidence based, are now fairly widely accepted. In view of the widespread, and justified, concern about a post-truth era, this seems encouraging, and worth investigating. A few examples

* In this post on equality of opportunity from 2003, I noted that “contrary to popular belief, there is less mobility between income classes in the United States than in European social democracies.” I was drawing on a 1999 book, The Real Worlds of Welfare Capitalism by Goodin, Headey Muffels and Dirven, which I’d reviewed a couple of years previously. In 2009, when I started work on Zombie Economics, I wrote about this again. However, I soon realised I was pushing at an open door. The decline of social mobility in the US had become part of the conventional wisdom.

* In 2004, some of the first studies of charter schools were coming out, showing that, contrary to the widely-shared expectations of education reformers, they weren’t showing any clear gains in student performance. I wrote about this fairly cautiously, noting that studies of this kind often fail to find any effect. As it turned out, however, the findings were replicated, particularly in the case of for-profit schools. This piece in the Washington Post (which used to be associated in some way with the for-profit testing industry, IIRC) shows how much the tide has turned against charters, and even more against for-profits.

* Here’s a post on minimum wages, drawing on the work of David Card and Alan Krueger (whose tragic death recently was a big loss to the economics profession). from the early 1990s. By then, the formerly orthodox view that minimum wages had big negative effects on employment was sufficiently out of favour to be revived in Slate (then famous, or notorious, for “contrarian” views that generally tended to support the establishment).

* Finally, I wrote a couple of mildly snarky pieces about the “Reading Wars” between phonics and whole language. This was one of the relatively rare cases in which the emerging evidence supported the cultural right. It’s pretty hard nowadays to find unequivocal supporters of whole language.

Looking at these examples, there’s a gap of about 10 years between the time the evidence emerged (or at least, emerged prominently enough for me to take notice) and the time the conventional wisdom adjusted. That doesn’t seem too bad. As the great replication crisis has shown, it’s unwise to take too much notice of an individual study on any social science topic.

Unsurprisingly, most of the examples above are cases where the emerging evidence was consistent with my broad political principles (I was never engaged in the Reading Wars, though I mostly lined up against the phonics advocates on other issues). I’d say that’s because most of the evidence we’ve had in the past twenty-five years or so has gone against the beliefs of the political right, who have had to retreat from the triumphalism of the early 1990s. But it’s obviously possible that there is confirmation bias at work. I’d be interested to see suggested examples of evidence shifting the conventional wisdom to the right in this period.