One favorite chestnut of the Left, is that if Bush was really supported by the people, he would enjoy better support than he does. Of course, since Bush's first election enjoyed a higher share of the Popular Vote than Clinton's first win, and Bush's second win enjoyed a higher share than Clinton's second win, this logically means that Bush is more legitimate than Clinton was, but never mind. Since Bush took a clear majority of the Popular Vote in 2004, the Left has to work hard to find a reason to discredit the win, even if it strains credulity.
Anyway, I was looking at the Popular Vote percentages in Presidential elections since 1948, and I noticed something about the share of the vote: Nineteen states and the District of Columbia have all given 51% of more of their Popular Vote to one candidate in at least the last three elections. Also, there are seven states plus D.C., which have given the losing candidate less than 40% of the Popular Vote for at least the last three elections straight. This means that there are some states which are pretty firmly Red or Blue, but it also means that there are thirty-one states which fluctuate from one election to another.
At first glance, this would appear to mean that those two facts are contradictory, but in actual practice, it means that each side, Republican and Democrat, has select states which they are almost certain to win or lose, before any specifics unique to the next election are considered. Next, the trend from recent elections and historical variances illuminates which of the so-called "battleground" states is truly up from grabs, and which carries a hidden loyalty. This characteristic, I must emphasize, is tied to party loyalty in greater measure than the charisma of an individual candidate.
As a result, each of the two major parties enters the General Election phase of their campaign with a solid expectation of its position, and that position is now almost always in the 45-55% Popular Vote range. The issues specific to a key state may make the difference in the election, and half a year before the ballots are cast, both sides know what will decide the election, if they have done their homework. Ironically, the best evidence indicates that the losing side miscalculated on that critical point.
What this means in summary, is not that the election does not matter, but that the mechanics of last-month tactics can be critically important to the outcome, provided the strategic decisions have understood the stress points which direct the outcome. Unless it's a charismatic nominee running against a dull moron, it will always be close.