The information below is from a reader comment, but I thought it highly enlightening, and worthy of main-page exposure:

(demtom): For reasons unknown, people persist in believing high war ratings provide an electoral boost for the president's party. There is NO evidence to support this, at least since the Spanish American war. Consider this:

1918 -- Election Day comes a few days before the Armistice. Democrats lose Congress. Two years later, they lose the presidency.

1942 -- 11 months after Pearl Harbor, FDR's numbers are in Shrub-range. Dems get clobbered in the midterms, losing 50 seats in the House and c. double-digits in the Senate (Arthur Schlesinger pointed this out last Fall, urging those suddenly dropping out Dem candidates to reconsider)

1944 -- Probably the best wartime election of the 20th century, but 1) the timing was perfect (war still on but victory more or less assured); 2) Dems made only minor gains, compared to the losses of '42; and 3) FDR was re-elected by his SMALLEST margin ever.

1946 -- First post-war election: Dems lose both houses.

And those were our big victories.

1952 -- Korea stalemate; Ike beats Adlai, and Dems lose both houses.

1966 -- Vietnam just underway. Dems lose more House seats than they'd won in the Johnson landslide. Two years later, they lose the presidency.

1992 -- We know what happened after the great Gulf War victory.

And I assume most people know what happened to Churchill within weeks of VE Day.

Why do people cling to this "war trumps everything" delusion? Was Wag the Dog that influential a movie?