[KarlInOhio]

I'm still looking for it, but I found this on google:

Release: May 9, 2000

UI political scientist says Democrats will win White House in 'near landslide'

IOWA CITY, Iowa -- Contrary to nearly every current political opinion poll, University of Iowa Political Science Professor Michael Lewis-Beck says Al Gore, the presumptive Democratic presidential candidate, will win the November election with 56.2 percent of the popular vote. "It's not even going to be close," he said.

http://www.uiowa.edu/~ournews/2000/may/0509forecast.html

[JLS]

"A less scientific, yet historically accurate model supports the prediction of a Democratic win, Lewis-Beck said, making him even more confident that he is right and the polls are wrong. In every post-World War II presidential election, the incumbent's party has won if the unemployment rate is lower in June than it is in January of the election year. That is almost certain to be the case this year, with unemployment at its lowest levels ever and still falling, he said."
___________________________________________________________________

I wonder what in the world these people are talking about:

Year Race UrateJan UrateJun Winner
1948 Truman-Dewey 3.4 3.6 Yet incumbent Truman won???
1952 Ike-Stephenson 3.2 3.0 Yet nonincumbent GOP wins????
1956 Ike-Stephenson 4.0 4.3 Yet incumbent GOP wins???
1960 Nixon-Kennedy 5.2 5.4 Incumbent GOP loses as predicted
1964 Goldwater-Johnson 5.6 5.2 Incumbent Dem wins as predicted
1968 Nixon-Humphrey 3.7 3.7 Nonincumbent wins despite no change in unemployment rate.
1972 Nixon-McGovern 5.8 5.7 Incumbent wins as predicted
1976 Ford-Carter 7.9 7.6 Yet Incumbent GOP loses????
1980 Reagan-Carter 6.3 7.6 Nonincumbent GOP wins as predicted.
1984 Reagan-Modale 8.0 7.2 Incumbent GOP wins as predicted.
1988 GHW Bush-Duke 5.7 5.4 Incumbent GOP wins as predicted.
1992 GHW Bush-Carter 7.3 7.8 Incumbent GOP loses as predicted.
1996 Dole-Carter 5.6 5.3 Incumbent Dem wins as predicted.
2000 GW Bush-Gore 4.0 4.0 Incumbent Dem loses with no change in the unemployment rate.

So it seems this model was wrong 4 out 14 cases. But three of them were the first three. Perhaps he was quoting this test from memory and did not remember that it worked for all but one election since 1960, not since WWII? And I do not see why he thinks the unemployment rate will for sure not fall from Jan? It was 5.6 in January. It is now 5.7. It could easily be 5.5 in June.