Wampum

One of the joys of blogging is that there are no timetables for posting. One is free to explore wherever one’s curiosity leads even if it takes a while. Recently, my curiosity led me to look at a bunch of vote totals for elections to the United States Senate.

It started when John Podhoretz, writing at The Corner wrote a most silly post. In its entirety, it said:

HARRY REID AND BUSH THE LOSER

Harry Reid vote total in 2004 = 490,232
Bush vote total in 2004 = 62,040,606

That is reminiscent of the maps and tee-shirts that some folks came out with after the 2000 election. The purpose of the maps was to show that President Bush actually won in a landslide over Al Gore. The maps proved their point by showing that the counties won by Bush had far more acreage than those won by Gore. There is no doubt about it, in terms of acreage, it was a landslide.

So, Podhoretz pointed out that a Presidential candidate drew more votes in fifty states than a Senate candidate did in one (small) state. Who would have guessed?

That would have been that except that I wondered whether Reid had run ahead or behind Bush in Nevada, the only place where both were on the ballot. I looked up the result and found:

Votes In Nevada 2004
George W. Bush 418,690 51%

Harry Reid 494,805 61%

Reid ran substantially ahead of Bush. While I was looking up the election returns for Nevada, another question tickled my curiosity. Everyone knows that 2004 was a disaster for Democrats, particularly in the Senate. They lost four Senate seats as Republicans increased their caucus from 51 to 55 seats. I wondered whether or not Republican Senate candidates, as a whole, ran ahead of or behind President Bush.

Mr. Bush got about 51.25% of the major party vote in 2004. Did Republican Senate candidates do better or worse?

I got my data here.

It turns out that Republican Senate candidates not only did far worse than the President, they did far worse than the Democrats.

Votes For Senators in 2004

Democrats 42,370,726 51.49%

Republicans 39,920,562 48.51%

Democratic Senate candidates ran ahead of Republican candidates by about the same margin as Mr. Bush ran ahead of John Kerry. Despite that vote count, Democrats got massacred in the Senate.

There were 34 Senate races in 2004. The GOP won 19 or about 55% of the seats while getting only about 48.5% of the total vote.

That seemed odd to me. Perhaps it was a fluke. I looked at the same data for prior election years. In 2002, Republicans won 22 of 34 races or about 65%. In cumulative votes, Republicans led 22,198,747 to 20,470,371. That is, Republicans won 65% of the races while garnering 52% of the vote.

In 2000, Democrats received 36,788,222 Senate votes while Republicans got 36,729,792. Despite getting essentially 50% of the Senate votes, Republicans won only 15 of 33 races or 45%.

Over the six year cycle, Republicans won 56 of the 101 Senate races while the Democrats won 45 (the discrepancies in the numbers you are now noting come from the combination of the special election in Missouri for the Carnahan seat, or the Talent seat if you prefer, and the party switch of Jim Jeffords). Over that same period, the vote totals are as follows:

Democrats 99,629,319 50.2%

Republicans 98,849,101 49.8%

Despite having received a (very slight) majority of the two party Senate vote over the last three cycles, Democrats are at a distinct disadvantage in Senate seats.

Please do not misunderstand me. I am not arguing that Democrats deserve control of the Senate by virtue of receiving more Senate votes. I am not arguing that it is unfair for the Democrats to have fewer seats than the national popular vote count would suggest. I am not arguing anything, really. We do not choose a Senate by adding up the national vote totals any more than we choose a President by adding up the acreage of the counties carried. It just seems odd that party vote totals are such a poor predictor of Senate seats.

There are a number of possible reasons for the discrepancy and I will look at one of them, the small state distortion, soon.