In the comments, some have asked whether or not this applies to the population of states represented by Democratic and Republican Senators as well. It does. The 45 Democratic Senators represent 296,673,557 people, while the 55 Republican Senators represent 289,532,423 people, according to the July 1st, 2004 population estimates by the census bureau. Another way of looking at it is that 174,919,609 people are represented by at least one Democratic Senator, and 171,349,042 people are represented by at least one Republican Senator. So, pretty much anyway you look at it, Biden is right. Democrats in the Senate represent the majority.—Chris

Today on C-SPAN, Senator Joe Biden is making/has just made a speech on the proposed rule change on filibusters of judicial nominations - the "nuclear option." He made an interesting statement, that "we [i.e. Democrats] represent the majority of the American people." This is interesting and true. 51% of voters since 2000 have voted for Democrats, 49% for Republicans.

One way to calculate this is to add up the total number of votes cast for Democratic and Republican Senate candidates for the last three elections, a period of time in which all 100 Senate seats have come up for election once. That total is:

Dem 100,494,879 (51%)
Rep  97,276,977 (49%)

Here are details of the calculation:
http://clerk.house.gov/members/electionInfo/2004/Table.htm
http://clerk.house.gov/members/electionInfo/2002/Table.htm
http://clerk.house.gov/members/electionInfo/2000/Table.htm

Year         R                       D
2004  39920562        44754618
2002  20630984        18959386
2000  36725431        36780875

I have been looking at the party-ID numbers in the Gallup data. I find evidence that party ID is not fixed over time. Here is a draft of the general argument.

Some of you may be familiar with my meta-analysis of state polls. This is a calculation in which I use the last three available polls in close states to calculate the probability of a Kerry win there. I then go through every possible combination of states to calculate the overall probability of winning the electoral college. A number of election prognosticators follow a similar procedure. My current result is Kerry 279 EV, Bush 259 EV, which is much more encouraging than national opinion polls.

Following the history over time reveals something interesting: