Whenever the party in power takes a hit in the midterms, it takes refuge in the past. Since the governing party almost always loses seats in off-year races, the president is said merely to have fallen prey to the ineluctable tides of history—much in the way that presidents who face economic depressions disown any blame by fingering the all-powerful "business cycle."
In particular, parties that incur setbacks in their presidents' second terms like to hide behind the "six-year itch," to use an ungainly term favored by political scientists. Typically a president's second-term off-year losses outstrip his first-term losses, and it's tempting to imagine some iron law of history at work—a structural force that has afflicted even popular presidents, such as Dwight Eisenhower in 1958 and Ronald Reagan in 1986. George Will, for one, made this argument on ABC last night.
But plans to invoke the six-year itch ought to be scratched. Politics has no iron laws: As circumstances change, so does political behavior. (Significantly, Bush in 2002 and Bill Clinton in 1998 both defied the trends, suggesting that gerrymandering, microtargeting, polarization, or other factors have scrambled historical patterns.) What's more, there have been too few sixth-year elections to be statistically meaningful. Most important, the variables in any given election—wars, recessions, scandals, social crises—matter more than tendencies built in to the system. On inspection, the six-year itch resembles less a chronic disease than a phantom illness on the order of chronic fatigue syndrome.
To be sure, certain structural forces do favor the nonpresidential party in the midterms. According to a theory of "surge and decline," presidents have coattails when they're elected, carrying into office their party-mates. But (to shift metaphors) when those legislators have to run a play without the president as their offensive line, many get thrown for a loss. Ronald Reagan's sixth-year setbacks fit this pattern: The Republicans gained control of the Senate in Reagan's 1980 landslide but lost it—despite the absence of scandal, recession, or other disaster—in 1986.
More provocatively, legal scholar Akhil Amar has suggested in America's Constitution: A Biography that the 22nd Amendment limiting presidents to two terms, ratified in 1951, has weakened second-term presidents. It's well-known that second-term presidents become embroiled in scandal: Nixon with Watergate, Reagan with Iran-Contra, Clinton with Lewinsky, Bush with faulty prewar intelligence (among other issues). One explanation is that re-elected presidents grow arrogant and reckless, as Nixon certainly did. But Amar suggests another reason: The 22nd Amendment effectively makes second-termers four-year lame ducks, unable to exact retribution upon antagonists on the Hill—thereby encouraging congressional investigations and commanding less party loyalty. By the same logic, the president is arguably now less able to pass legislation and otherwise work his will in his sixth year, thus deepening his midterm losses.
Nonetheless, in almost every case of a six-year itch, Occam's Razor suggests more direct reasons for a president's party's losses. In 1874, for example, the Democrats made big gains in Republican President Ulysses Grant's second term. Yet they plainly benefited from the financial panic of 1873, as well as from the Credit Mobilier scandal—considered the third-worst presidential scandal after Watergate and Teapot Dome.
In 1938, the six-year itch did seem to be at work in the defeats that the Democrats endured after scoring routs not only in 1932 and 1936 but also—bucking historical trends—in the off-year contests of 1934. Yet again, more proximate causes abound. Franklin Roosevelt blundered in 1937 when he proposed a massive overhaul of the Supreme Court to get it to uphold his New Deal legislation. Moreover, after much progress in reducing joblessness and reviving public confidence, he cut government spending to bring the budget into balance, thereby kicking the economy back into recession. By 1938, the unemployed had swelled from 5 million to 12 million, crippling the Democrats in November.
Even in 1958, when the widely beloved Eisenhower lost 48 House and 13 Senate seats, contemporary events, more than structural dynamics, were at fault. Most significantly, the severe economic downturn that year hit especially hard in the Midwest, depressing turnout among Eisenhower's most natural constituents. Besides, the Soviet Union had launched just Sputnik, spurring a panic about the state of American defense, education, and—most important—nerve and morale. This disillusionment with Ike's governance was an early expression of the now-canonical critique of his presidency—what the journalist William Shannon termed "The Great Postponement."
A similar disillusionment may have been in effect his year, as historian Niall Ferguson argued before the election. There's also a whiff of 1966 in the air, as I've noted elsewhere—1966 was another year in which voters rebelled (in a second-year, not a sixth-year, election) against one-party rule. This year, the discontent that exit polls found with what they labeled "corruption" should be more properly understood as a rebuke to the general arrogance of the unchecked Republican Party. Think of corruption in Lord Acton's sense.
In retrospect, though, the Republicans' losses seem most similar to those in three other midterm races: 1918, 1950 (not properly a sixth-year midterm, since Harry Truman entered the presidency in 1945 through accession, not election), and 1942 (technically a "10th-year" itch). All of those setbacks came amid wars. In 1950, Republicans painted the controversial adventure in Korea as a result of Truman's weakness. In 1942, Democrats suffered when World War II was going poorly in both the Pacific and European theaters. In 1918, the Allies were nearing victory in World War I, but there was much resentment in the land, especially after Woodrow Wilson, having promised to keep the United States out of war, now explicitly called on Americans to vote Democratic as a show of support for his policies.
Wars help presidents so long as the rally-round-the-flag effect holds up. The Iraq war did so for Bush in 2002 and even 2004 (though by then it was becoming uncertain whether the Iraq war was helping or hurting Bush). On the other hand, a conflict that has no clear end in sight vexes Americans of all political stripes, summoning up deep strains of both conservative isolationism and liberal anti-imperialism. As my Rutgers colleague Ross K. Baker, a congressional expert, wrote last spring, "Combat fatigue is not a condition found only on the battlefield; it is also an affliction that has often been diagnosed in the voting booth." If there's a history lesson to be drawn from this year's election results, that one would be closest to the mark.