I took a class from Fair, I don't put too much stake in what he says here.
There are 2 flaws to his method.
1) He uses only economic data to the exclusion of everything else. This is understandable given his bias as an economist, but elections don't only hinge on economics.
2) He uses actual economic data, when often elections are determined by public perception of the economy. Which generally trails reality by quite a few months. This is why Fair's method predicted that GHWB would beat Clinton. Because although the economy had turned around before Nov 1992, the public didn't recognize the improvement til after.
"This is why Fair's method predicted that GHWB would beat Clinton. Because although the economy had turned around before Nov 1992, the public didn't recognize the improvement til after."
-- No. He mispredicted that one due to the Perot Factor. He got 18% of the popular vote and up to 24% in some key states, including 23% in CA (Bush won CA in '88). If Perot didn't run GHWB would have won, bottom line. Furthermore, Fair's model would have still been accurate.