This week I write my last two columns for the year, so it's probably natural to be a bit reflective after such a tumultuous election year.
Those who tried to impose a Tip O' Neill-style "all politics is local" template on the 2006 elections missed the boat. This was not a normal election like 1996, 1998, 2000, 2002 or 2004 were; it was more like the elections in 1958, 1966, 1974, 1982 or 1994. But just as those who insistently tried to view the 2006 election through a normal election prism missed the boat, in all probability, anyone trying to preview the 2008 election through the 2006 lens will almost certainly be just as wrong.
Sure, Democrats may screw up and lose their House or Senate majorities in 2008, or one party may nominate an unusually strong or weak candidate that will alter normal presidential election year turnout patterns and tilt the playing field in favor of one party or the other. But the odds of 2008 becoming a major change election are pretty slim. It would be a lot safer to go back to the playbook of 1996 through 2004, when individual campaigns and candidates mattered more than they did in 2006 and when there wasn't an invisible hand pushing the candidates of one party forward like you saw in 1994 and 2006.
Historically, the incumbent re-election rate for the House is around 94 or 95 percent. This year it was 94.3 percent and the only oddity was that every incumbent who lost was from the same party. But this 94 percent re-election rate is in stark contrast to the 98-plus percent re-election rates that we had in the five previous elections, the longest string that high in the post-World War II era. My hunch is that we won't see 2008 in that range, and may not see one like that for a while. The sight of seeing so many incumbents lose this year will not be lost on would-be challengers, whether they are Republicans or Democrats.
After a couple of 98-plus percent re-election years, it became self-fulfilling. More potentially strong challengers were discouraged about running because they saw little chance of winning. Weaker challengers in normal years almost inevitably result in fewer incumbents losing. And on it goes, year after year, until the political environment becomes so toxic for one party that the cycle is interrupted, as it was in 2006.
One blessing for Republicans this year was that so few strong Democratic challengers won. To be sure, it wasn't for a lack of trying. Several times I have written in this column that Democratic Congressional Campaign Committee Chairman Rahm Emanuel and his team did everything short of kidnapping the children of prospective candidates -- the proven winners, the state senators and representatives, county executives and prosecutors, and people that have run and won before and know what it takes to win a tough campaign. In case after case, those proven, battle-tested Democrats opted not to run, leaving the Democratic nominations to go to others who often had never run for office before, or had never won, or had won but had not had tough races. As it turned out, the year and the climate were so good that many won anyway, but Republicans should be grateful that Democrats had so many novice and uneven candidates or it could have been worse.
Just look at Michigan, where not one incumbent in either party had a serious general election opponent. Or look at how many Republican incumbents in Ohio -- in districts that President Bush won by three points or less in 2004 -- had no serious opponents.
In short, for challengers in both parties, there is blood in the water. My hunch is that we will have more incumbents in both parties with formidable opponents in 2008 than we have seen in some time. At the same time, keep in mind that in normal years, true upsets are fewer and further between than we saw this year. The action typically takes place in districts where the presidential election results were reasonably close.
Using our (PVI), which indicates how much more Republican or Democrat each district voted in the 2000 and 2004 presidential elections compared with the national results, most of the real action -- incumbent or open seats changing hands -- takes place in the districts that are no more Republican than R+3 (meaning that district voted 3 percentage points more Republican in 2000 and 2004 than the national major party vote) to D+3 (3 points more Democratic than the country).
In short, it's safer to assume that any given election will be relatively normal (all politics is local), but always keep an eye on national polling diagnostic indicators like right direction/wrong track, Congress' and presidential approval ratings and the generic congressional ballot test (keeping in mind the 5-point, pro-Democratic skew) for any signs that an upcoming election might be the abnormal kind -- the nationalized election when Tip O'Neill's "all politics is local" adage simply isn't the case.