There has been an increasing about of back-and-forth about the meaning of victories in various primary and caucus states. Do blue states matter more than red states? Do a couple of big states matter more than a large number of little states?
These arguments, of course, are mostly just static to fill the nervous void until the next wave of primary results comes along. After all, the Democratic Party has come up with an ingenuous way to keep score: they are called delegates, and they are allocated in such a way that they will essentially reflect the relative share of the popular vote in all officially-sanctioned contests around the country.
However, there are other respects in which some states do potentially matter more than others. One is in the electoral math for the general election. If one candidate were performing systematically better in swing states than another, that might be an argument to nominate them. But both candidates have had their share of successes and failures in these contests -- Obama winning Missouri, Iowa and Colorado; Clinton winning Nevada and Florida -- so it is hard to make a case one way or another on these states.
Another set of important states, however, has to do with the locations in which we are liable to have the most competitive Senate races. And these states obey a different set of rules. For one thing, Senate seats are of course allocated two men (or women) per state, regardless of population size. So in the case of the Senate, a lot of little states really do count more than a couple of big states. For another thing, this political cycle will be unique in that we are on offense, and that Senate contests will take place almost entirely on red and purple turf.
These contingencies would seem to favor Obama, with his appeal across a wider range of states, and particularly in red and purple states. And in fact, Obama has done overwhelmingly better than Clinton in places where we can expect to have a competitive Senate race in 2008.
The Cook Political Report currently projects that 12 Senate races will be at least reasonably competitive: 10 in which we're playing offense, and 2 in which we're playing defense. Certainly, some people here would like to add a couple of other states like Idaho to that list, but we'll go with Cook's list for the time being.
Here are those 12 states. You should begin to detect a pattern here relatively quickly.
- Alaska. Obama won overwhelmingly.
- Colorado. Obama won overwhelmingly.
- Louisiana. Obama won overwhelmingly.
- Maine. Obama won overwhelmingly.
- Minnesota. Obama won overwhelmingly.
- Mississippi. No contest yet; Obama favored.
- Nebraska. Obama won overwhelmingly.
- New Hampshire. Clinton won narrowly.
- New Mexico. Still unresolved; Clinton likely to win narrowly.
- Oregon. No contest yet; Obama favored.
- South Dakota. No contest yet; Obama favored.
- Virginia. No contest yet; Obama favored.
Obama has already won 6 out of the 12 contests by double digits, and is the heavy favorite to win four others. Clinton narrowly won New Hampshire, and will probably hang on to win New Mexico. To get a bit more quantitative about this...
In the eight states so far that have held primaries or caucuses, Obama has won by an average of 24 points. If we make simple projections of the results of the four remaining contests based on current polling averages or the results in neighboring states, we see that he's likely to win the entire group of 12 states by at least 20 points on average.
How will winning these states by 20+ points in the primaries translate in terms of greater coattails in November? I don't know, but I can give you some reasonable idea of the importance of said coattails. The chart that follows indicates the average number of Senate contests that were within a given percentage point margin between the last four election cycles (2000, 2002, 2004, 2006).
On average, for example, 3.75 Senate contests per cycle have been resolved by 3 points or fewer, 6 have been resolved by 5 points or fewer, and 9.5 have been resolved by 10 points or fewer. These results actually follow a pretty good rule of thumb: in any given year, X number of Senate contests are decided by X number of percentage points. For example, on average each cycle, 1 election comes within 1 percentage point, 7 elections come within 7 percentage points, 12 elections come within 12 percentage points, and so forth. So, if Obama's coattails are worth an additional 3% of the vote as compared with Hillary's, we can expect 3 additional Senate seats to change hands; if they're worth an additional 5% of the vote, we can expect 5 additional Senate seats to change hands, and so forth. This is a pretty important advantage considering how important each Senate seat is. Some of you might have weak preferences between Obama and Clinton, but you'd probably prefer Obama with a 57-seat majority in the Senate to Clinton with a 55-seat majority, or Obama with a 60-seat majority to Clinton with a 56-seat majority.
Let me conclude with a graphic that gives you a pretty good idea of Obama's 50-state strength, as well as the state of the delegate race so far.
The circles correspond to the size of the net delegate gain from each state that has voted so far. The chart should give you a sense that Obama has actually developed a fairly meaningful delegate lead. But more importantly than that, that this lead has come from big wins in almost literally every corner of the country. The key Senate races this year will occur in far-flung places like Alaska and Maine and Nebraska, and that plays perfectly into Obama's strategy.