Regardless of the results, we should use caution in interpreting this year’s Senate elections as reflecting the views of the entire nation. Two aspects of the U.S. Constitution’s rules on Senate elections make it possible that the political leanings of the Senate seats up in any particular year will be unrepresentative of the nation as a whole.
The first is by design: two Senate seats are allotted to each state, regardless of population. Americans living in Wyoming (the least populous state) have substantially more influence in the Senate than those living in heavyweight states such as California or Texas (which have 66 times and 45 times more people, respectively). Thus if one party draws its national strength from large-population states, in any given year it is more likely to face an unrepresentative number of tough Senate elections in smaller states. Currently, that party is the Democrats. The population of the median state won by Barack Obama in the 2012 presidential election is about 5.5 million; the median state won by Mitt Romney has a population of 3.4 million.
The second reason for the unrepresentativeness of a cycle’s Senate elections is more of a matter of historical circumstance. The Constitution directed that senators be assigned equally into three classes (Class 1, Class 2 and Class 3), with the six-year terms for each class to be staggered by two years. This means that Senate elections occur in any particular year in only about two-thirds of the 50 states. (In 2014, elections are for the seats in Class 2.) The initial assignment of senators to classes (undertaken by the Senate shortly after it first convened in 1789) appears to have been carried out with an eye toward balancing the young nation’s population, regions and political views among the three classes. But that has changed over time. As this great post by the University of Virginia’s Geoffrey Skelley shows, states with Class 2 seats now make up a much smaller share of the national population (52 percent) than do those with Class 1 or Class 3 seats (at 75 and 73 percent, respectively). And as many observers (including Ben Highton here on the Monkey Cage) have noted, states with Class 2 seats are much more Republican-leaning than the nation as a whole this year.
Taken together, the rules on seat allotments and classes have yielded a Senate election cycle in 2014 that is profoundly unrepresentative of the nation as whole — and particularly tough for Democrats. A good measure of the parties’ relative strength in the states holding Senate elections is the share of the state vote each earned in the most recent presidential election. The figure above plots (in blue) the Democratic share of the two-party presidential vote in the median state of those holding Senate elections from 1950 through 2014. For comparison, it also displays (in gray) the share of the national popular vote the Democrats received in the most recent presidential election.
For most of the past six decades, these two trends tracked each other very closely: the parties’ relative strength in the set of Senate seats up for election was no different from their strength nationally. But that changed after the 2000 presidential election, in which the Republican Party’s dominance in the South emerged in full force. This in turn led Class 2 Senate seats to be particularly strong for the GOP: There is a Class 2 seat in all but one state of the former Confederacy — Florida.
Senate elections took place in Class 2 states in 2002, 2008 and now 2014. In each of these cycles, there is a substantial difference between the parties’ relatively even performance at the national level and the Republicans’ strength in Senate seats up for election. This year, the gap is profound: in the typical Senate election state, Republican presidential nominee Mitt Romney ran ahead of his national margin by more than seven percentage points.
Simply put, this year’s Senate elections are unrepresentative of the nation to an extent that is unprecedented in elections held in the post-war era. So when we begin to sift through the results on Election Night, the number of Senate seats won and lost will tell us less than we might like about where the two parties stand in the minds of American voters.