As expected, today Senate supporters of the Marriage Protection Amendment, which would change the US Constitution to ban same-sex marriage, failed to muster the votes needed to end debate — and fell far short of the two-thirds vote that would have been needed for passage. The vote was 49-48, slightly better than the 48-50 on an identically worded amendment two years ago (what a coincidence that these votes come up only in election years!).
But the Republicans have gained 5 seats in the Senate since the 2004 vote, so why was there only 1 more vote in favor of the amendment? I decided to get into the details and see exactly what happened.
First, 84 senators voted the same way both times, so we can ignore them. Then, there were some changes as a result of the 2004 election:
- In New Jersey and Oklahoma, senators were replaced by senators of the same party who voted the same way. (net change 0)
- In Colorado, Campbell (R) was replaced by Salazar (D), who still voted nay. Similarly, in Georgia, Miller (D) was replaced by Isakson (R), who still voted yea. In each case, a senator who voted against the majority of his party was replaced by a senator from the other party who voted the same way. (0)
- In Illinois, Fitzgerald (R), who voted yea, was replaced by Obama (D), who voted nay. (-2)
- In North Carolina, Edwards (D), who didn’t vote, was replaced by Burr (R), who voted yea. (+1)
- The big one: In Florida, Louisiana, South Carolina, and South Dakota, Democrats who voted nay were replaced by Republicans who voted yea. (+8)
So the 2004 elections gave Republicans a gain of +7 on the vote. Why was their actual improvement only +3? The answer lies in what happened with those who were in the Senate for both votes but acted differently in the two years:
- In Connecticut and West Virginia, Dodd (D) and Rockefeller (D) voted nay in 2004 but didn’t vote this time. (+2)
- In Massachusetts, Kerry (D) didn’t vote in 2004 (when he was running for president) but voted nay this time. (-1)
- In Nebraska, Hagel (R) voted yea in 2004 but didn’t vote this time. (-1)
- In New Hampshire and Pennsylvania, Gregg (R) and Specter (R) voted yea in 2004 (when both were up for reelection) but voted nay this time. (-4)
The difference is explained by the only two senators who actually changed their votes. Gregg and Specter, both Republicans, went from supporting the amendment to opposing it once it was no longer an election year for them.
Is there a standardized digital format for these bills that you could use to do a diff across years and generate these sorts of comparisons automatically? It’d be particularly easy for the senate, because the number stays the same.
—David Eads • 2:11 pm
That’s more or less what I did to start with: wrote a Perl script to parse the HTML of the vote pages. The format is consistent, but of course there’s no guarantee that some designer won’t change it tomorrow. The House had a project to provide information in XML but it looks like it hasn’t been doing votes this year (the “Current Congress” link goes to 2005). I haven’t been able to find anything similar for the Senate.
—Keith • 2:38 pm
Hi Keith: Thanks for your work, but wasn’t the vote this year 48-49, instead of 49-48 as you have it? As distinct from the 2004 vote of 50-48? I’m trying to reconcile your analysis with this peice in Forbes, http://www.forbes.com/business/feeds/ap/2006/06/07/ap2799777.html which attributes the significance to Judd, R-NH, changed, and Hagel R-Neb, absent, and my brain doesn’t work in the mornings.
It’s always nice to know your effort’s noticed, even if it’s by a complaint, yes? Best Regards, Steven
—s • 9:41 am, June 9
Thanks, Steven. I’m going by the votes reported on the Senate site (linked above from the vote counts). And the AP article in Forbes does say 49-48, as I do. I don’t understand why it has 50-48 for the 2004 vote, or why it omits Specter’s changed vote. The Post article also used to have a confusing sentence saying the vote was slightly worse for the Republicans this year, but that’s been removed now. I think some reporters just got some bad information early on and it spread and hasn’t been corrected everywhere.
—Keith • 9:56 am, June 9