The Bradley Effect that Never Was [LINK]
My latest letter to the Globe:
The Globe is right to identify the "Bradley effect" as a hypothesis, not as a settled reality that has now been overturned. Named after former Democratic Mayor Tom Bradley of Los Angeles, who twice failed in California gubernatorial campaigns, the idea was that white voters would falsely tell pollsters they intended to vote for an African American candidate in order to conceal their latent racism.
Readers may be interested in the analysis of Lance Tarrance, the pollster relied upon by Bradley's Republican opponent. Tarrance insists there never was such an effect, and that the theory was floated by a rival polling firm that predicted a Bradley win after badly misinterpreting its data.
The 2008 election is unquestionably an historic landmark and a source of great pride. But relying on mythology, it's easy to overestimate the underlying change in attitudes it actually represents. I have no reason to believe Americans had been predisposed to vote against African-American candidates in the recent years prior to Obama's extraordinary campaign, any more than they were unwilling to vote for a woman. This is a good thing all around.