by Anthony G. Greenwald, professor of psychology, University of Washington

and Bethany Albertson, assistant professor of political science, University of Washington

Sifting through overnight results after the Super Tuesday primaries, we have found that race still plays a role in American politics but that it showed up in surprising ways in the tallies from states holding Democratic primary elections so far this year.

Early analysis of primary counts and polling data from the final week of the campaign indicated that pre-election polls exaggerated support for Sen. Barack Obama in two states with relatively low black populations –California and Massachusetts. But the reverse was true in Alabama and Georgia, where blacks make up a larger bloc of voters. The same phenomenon is seen in the earlier primary states of New Hampshire and South Carolina.

The findings in Alabama and Georgia suggest the discovery of a new “reverse” Bradley effect.

The Bradley effect was first noticed by survey researchers in 1982 when black Los Angeles Mayor Tom Bradley had a solid lead in the pre-election gubernatorial polls, but lost a close election in California to his Republican opponent. Results from that and other races involving black candidates indicated that, for whatever reason, pre-election polling tended to overstate support for black candidates compared with their actual vote percentages.

In research we jointly undertook last December, we analyzed data from an online test that measures unconscious or automatic preferences. On the basis of our findings, we surmised that the Bradley effect might well repeat itself in 2008. Our more recent findings, however, suggest a more complicated pattern.

In our latest analysis, we had available not only late-polling data from 11 Super Tuesday states, but also earlier data from New Hampshire and South Carolina. As shown in the graph, New Hampshire was another state that showed the Bradley Effect, while South Carolina showed the Reverse Bradley Effect.

Actual results of the Democratic primaries in California, Massachusetts, New Hampshire, South Carolina, Georgia and Alabama compared against late polling numbers all substantially exceeded the polls’ expected margin of error, being off by between 8 and 18 percentage points. By contrast, we found that Republican vote totals and poll numbers were off substantially only in Massachusetts where Mitt Romney’s winning margin was less than predicted.

As the 2008 election season plays out, we plan to investigate further the possible Bradley and reverse Bradley effects.