SOME pundits are saying that voters are likely to succumb to the so-called six-year itch for tomorrow's midterm elections. That's the urge to scratch the president's party in congressional elections in the sixth year of his administration.
So the Democrats could lose big. That is the historical flow. No doubt about it. But I think you can make too much of it. I think those who subscribe to the six-year itch theory have a point, but they overlook or ignore something, which leads to exaggeration.
Let's check the record. Strictly interpreted, there have been only four six-year itch elections since the people (instead of state legislatures) began to choose senators in 1914. Those were: 1918, which was Woodrow Wilson's sixth year in the White House; 1938, in Franklin D. Roosevelt's sixth year; 1958, in Dwight Eisenhower's sixth; and 1986, in Ronald Reagan's.
Wilson's Democrats lost 19 seats in the House and six in the Senate in 1918. Roosevelt's Democrats lost 71 seats in the House, six in the Senate in 1938. Eisenhower's Republicans lost 47 House seats, 13 in the Senate in 1958. And Mr. Reagan's Republicans lost five House seats, eight Senate seats, in 1986.
Those four elections work out to an average loss by the president's party of 35 House seats and eight Senate seats.
One explanation for the six-year itch theory is that voters who lean to the president and his party have become bored with him and his personnel and his policies, and they take it out on his party in Congress by staying home on Election Day.
Some political analysts go further and designate any election in which a party took possession of the White House six years before, and kept it, as an "itch" election, regardless of whether the same president was in the office for all six of those years.
By that definition, the 1974 election also qualifies. That was the year Richard Nixon, elected in 1968 and re-elected in 1972, resigned to avoid impeachment. His successor was Vice President Gerald Ford. In 1974, Democrats gained 43 House seats and three Senate seats. Add that in and the six-year itch average losses for a president's party rise to 37 in the House and drop to seven in the Senate.
By the broader definition, 1966 would be an "itch" election year. Then, Lyndon Johnson had only been president three years, but he had been the incumbent in the presidential election of 1964, having succeeded John F. Kennedy's in 1963.
Similarly, 1926 would be an "itch" year. Then, Calvin Coolidge had been elected president in 1924, when he was the incumbent as a result of being vice president when President Warren Harding died in 1923.
Johnson's Democrats lost 47 seats in the House and three in the Senate in 1966. Coolidge's Republicans lost 10 seats in the House and six in the Senate in 1926.
I would add one more election to the "itch" column, though it occurred in a year in which the president's party had been in power for more than six years. Harry Truman assumed the presidency upon Roosevelt's death in 1945, was re-elected in 1948, and, in 1950, his Democrats lost 29 House and six Senate seats.
Counting all eight of those six-year itch elections, the average loss for the president's party in the House has been 34 and six in the Senate.
That sounds pretty bad, but what is overlooked is that voters get a pretty strong itch to scratch after a president has been in office only two years, long before familiarity has produced boredom.
In the first midterm election of every elected president since 1914, the average House loss by his party was 27 seats, and the Senate loss was three. Of the four presidents who were still in office at the time of the sixth-year election, two (Wilson and Mr. Reagan) saw their party's congressional candidates do better than they had done four years before in the administration's first midterm voting.
Judged by the historical record, the six-year itch is more irritating to a president's party than the two-year itch, but not by that much.
Theo Lippman Jr. is a retired editorial writer for The Sun.
Pub Date: 11/02/98