January 31, 2006
Will the Republicans Suffer a Sixth Year Slump?
By Jay
Cost
In the course
of making electoral prognostications, political pundits this year
are still referencing what is known as the “sixth year slump”.
This is the slump from which the congressional party of the President
inevitably suffers. Well, almost inevitably. The trend was seriously
disrupted in 1998 – when the Democrats, despite sporting
a president who was having a very bad sixth year, actually gained
5 House seats. It was essentially gospel in the lead-up to 1998
that the Republicans would gain seats. But, of course, they did
not. And so, the importance of the sixth year slump as a predictive
tool declined. Today, it is usually only mentioned by columnists
at the end of a laundry list of reasons that the Republicans will
lose seats: “War in Iraq, uncertainty about the economy,
presidential job approval, Abramoff,…and, oh yeah!, sixth
year slump.”
It is interesting
that pundits are still making mention of it at all. One would
think that 1998 would have ended any references to it as an independent
causal factor. But alas, its use persists. It is definitely true
that parties of the President have tended to lose seats in off-year
congressional elections. However, these seat losses have nothing
to do with the fact that the President is in his sixth year. It
is a coincidence and therefore not worthy of mention as a reason
to expect anything this year. Midterm congressional elections
– in the aggregate – hinge upon presidential job approval
and economic performance. They do not hinge upon the position
of the President vis-à-vis the Twenty-Second Amendment.
The essential
phenomenon that supposedly constitutes the sixth year slump is
the following: the President, despite being two years from reelection,
loses seats in Congress. Sixth year slumps often follow sizeable
electoral victories. Truman lost twenty eight Democratic House
members in his sixth year. Eisenhower lost forty eight Republican
House members. Nixon/Ford lost the same number. Reagan lost five
House members (and control of the Senate). These losses have been
consistent across parties and across presidential popularity.
Eisenhower was extremely popular in 1958. Reagan was also very
well liked in 1986. Truman and Nixon/Ford were not liked as much.
It is against this trend that 1998 stands as an exception. The
Democrats actually gained five seats in the House that year.
The question,
therefore, is how do we explain this trend? Perhaps it would be
helpful to note that there seems to be a similar pattern in the
second year of a president’s term. Since Roosevelt, every
president except George W. Bush has suffered losses in the House
in the second year of his term. It is also worth noting that in
presidential election years – with the exceptions of George
W. Bush’s two elections and Bill Clinton’s reelection
– every president since Woodrow Wilson has brought with
him a significant number of his partisans to the House.
In other
words, there seems to be some kind of “surge and decline.”
Until recently, the congressional party of the next President
surges when he is up for reelection and declines when he is not.
This is a more thorough description of the pattern we see in congressional
elections than “sixth year slump”, but of course it
does not explain it. We still do not know why this pattern exists.
Pundits, for their part, rarely try to explain this (or any) electoral
pattern. They merely take it as a given. But, without a good theory
that explains the pattern, we really should not follow their lead.
It might be the sign of some deep cause inherent to American politics.
But it might also be coincidental to the run-of-the-mill electoral
process we see every year.
Political
scientists have, in the last two decades, transitioned from the
second perspective to the first and then back to the second –
where a firm consensus on the matter has developed. Many were
once inclined to accept the idea that something unique about congressional
midterm elections hurt the party of the President. Now, however,
most political scientists are inclined to reject this. They now
believe that the aforementioned variation can be understood without
recourse to some unique theory. In other words, the sixth year
slump is not real.
The causal theory behind surge and decline was best put forth
by James Campbell in his The Presidential Pulse of Congressional
Elections. According to Campbell, it is the differing composition
of on- and off-year electorates that creates the effect. The electorate
in a presidential election tends to be larger than an off-year
election. It also, so the theory goes, tends to be less partisan,
less informed, and therefore more susceptible to “short
term” political forces – like the campaign speeches,
debate performances, personality of the candidates, etc. Those
who vote just in a presidential election are the people that offer
successful presidential candidates their landslides. In so doing,
they also vote for members of the President’s congressional
party. In off-year elections, they do not come out to vote. Those
who remain are the ones who are more informed, more partisan,
and more likely to vote according to long-term forces like political
values, beliefs, etc. They vote the way they always vote, so the
President’s congressional party does not get the boost that
it received two years prior, and therefore it loses seats in Congress.
Campbell’s
theory is an interesting one. It is easy to understand and intuitively
plausible. It does a good job of explaining much of the data we
have discussed – the “coattails” the President
often provides upon election, and the harm that comes to his party
when he is not on the ballot. Unfortunately, this theory fails
to account for key pieces of data, as of yet unmentioned. Most
notably, it cannot explain the fact that midterm electorates and
presidential electorates tend to be equally partisan. Thus, adding
or subtracting presidential voters should not make much of a difference.
In other words, the two types of electorates are not different
enough for this theory. It also fails to account for the size
of congressional seat changes. Why, for instance, should Clinton
– who barely eked out 43% of the vote in 1992, lose fifty
four House seats in 1994? Why should Reagan, whose 1984 victory
stands as one of the largest in American history, only lose five
seats? Surge and decline offers no compelling answer. It cannot
account for the magnitude of the surge and decline. So, while
the pattern remains, this theory does not do a sufficient job
of explaining it.
What then,
explains the variation between on-year and off-year elections?
As it turns out, the above pattern of surge and decline is not
all that unique. If we brought forth presidential approval ratings
and variables that describe the economy for the Eisenhower, Nixon/Ford,
Reagan and Clinton administrations, we would see that they explain
the surge and decline quite well. As a matter of fact, recent
work suggests that upwards of 70% of the variation in midterm
elections is explained by job approval and economic performance.
In other words, there is actually no special pattern to explain
at all. The rise and fall of aggregate partisan fortunes in Congress
really depends upon nothing more than presidential popularity
and the state of the economy. In the past, these have tended to
cut against the President during off-year elections – to
claim this is very different than claiming that there is something
special about off-year elections that cuts against the President.
Surge and decline is, in other words, largely a coincidence.
So, the sixth
year slump is really quite ephemeral. It indicates no unique causal
process – there is nothing special that induces a president
in his sixth year to lose seats. The variation we see in on- and
off-year elections can all be explained by the regular seat changes
caused by changes in job approval and economic performance. That
is why Bill Clinton and George W. Bush bucked the trend in 1998
and 2002, respectively. They were popular and/or they were leading
the nation during a time of prosperity. Thus, there is no reason
to think that, simply because George W. Bush is in the sixth year
of his presidency, he will necessarily lose seats.
Jay
Cost, creator of the Horse
Race Blog, is a doctoral candidate of political science at
the University of Chicago. He can be reached at jay_cost@hotmail.com.
©
2000-2006 RealClearPolitics.com All Rights Reserved
