The recent spate of good news for President Bush (in Iraq, in a hotly contested special election in San Diego, and in the decision of the prosecutor not to press charges against Karl Rove in the CIA leak case) has prompted yet another reassessment of the prospects for Democrats this November of winning majority control of the House and/or Senate. Polls conducted subsequent to these events reported a slight up-tick in public evaluations of Bush and his handling of the war in Iraq, suggesting that his stunning political collapse since winning reelection may have finally bottomed out and begun to reverse. Just as likely, however, is that this minor bounce in the polls will prove ephemeral, as have previous ones following supposed turning points. Critical public views of the president, the war in Iraq, and the economy have hardened over the past year and are unlikely to change without more measurable and durable improvements in the underlying conditions. Are such improvements possible in the barely four months remaining before the election? Can a skillful national campaign frame the partisan choice for the electorate to compensate for public unhappiness with the president and his party? Even if not, does the current structure of competition in House and Senate elections insulate the majority party from a negative referendum on the administration sufficient to make a change of party control of Congress unlikely?
Uncertainty over what to expect in the midterm elections is based partly on a belief that critical changes in contemporary politics have rendered historical lessons inapposite. One of the few iron laws of American politics was that the president’s party loses House seats in midterm elections. (The pattern in the Senate was less consistent, depending importantly on which seats are up in the election cycle.) The only exception between the Civil War and the near end of the twentieth century was the 1934 election, in the midst of the last major party realignment. Elections in the sixth year of an administration (the “six year itch”) produced on average substantially larger seat losses. The size of the midterm loss was related to how many seats the president’s party gained in the previous election (the larger the gain, the more seats at risk), how the economy performed in the year preceding the election, and how the public evaluated the performance of the president. A modest House seat pickup (or loss) in the previous presidential election, a healthy economy, and a presidential approval rating of 50 percent or higher minimized the normal seat loss. Each of these factors plus a negative public reaction to Republican plans to impeach President Clinton allowed the Democrats to break the pattern and gain five seats in 1998. Four years later, the “iron law” was shattered again: with George W. Bush in the White House, Republicans picked up eight additional House seats. GOP fortunes were boosted by the fact that they lost House seats in the 2000 elections, the President enjoyed post-9/11 approval ratings well above 60 percent throughout 2002, and the Republican campaign effectively elevated terrorism over the economy as the central public concern.