How to screw up “analysis”, by MSNBC.

Barack ObamaNow that Obama is the Democratic frontrunner it’s time for the mainstream media to start launching against him. That font of liberal goodness, MSNBC, has added to the fray with a bogglingly dumb article attempting to explain how Barack Obama is partisan, by looking at his voting record.

Obama’s roll call votes in 2005, 2006 and 2007 have been analyzed by the non-partisan journal Congressional Quarterly which found him to be a Democratic Party loyalist.

CQ didn’t use every single roll call vote to make this determination, instead it analyzed votes where a majority of Democratic senators opposed a majority of Republican senators.

In 2007, Obama voted with his fellow Democrats 97 percent of the time. In 2006, his score was 96, and in 2005, he again netted a 97 percent rating.

Well… yeah. He’s a Democrat, and got elected because he represents Democratic ideals. It’s a little dumb to think he’d step vote with the Republicans just to “step across the aisle” and show a willingness to work with the other side. Actually that’s an extremely stupid metric because it doesn’t show how partisan he is, just where his political stance is.

A far more logical way to check his willingness to work with Republicans is to look at the bills he sponsors or co-sponsors. The Library of Congress’s website doesn’t like to store links, rather it makes you work through temporary search results, so I can’t link to this directly.

In the 109th Congress, Barack Obama sponsored 152 bills. Of those 152, 96 had a co-sponsor. Of those 96, 41 had at least one Republican as a co-sponsor, often multiple. So depending on how you look at it, Obama reached across the aisle either 27% or 43% of the time when he sponsored a bill.

I had Will check out the 110th and come at it from the other side, just looking at the first 200 bills on which Obama was a co-sponsor. The number wasn’t as good there, but still sat at a solid 19 (just under 10% this time). Assuming there are no spikes or drops, we’ll call that the standard number.

So when Obama sponsors a bill, there’s between a 1 in 5 and a 1 in 2 chance a Republican will come with him, and when Obama signs on as a co-sponsor, there’s a 1 in 10 chance it will be with a Republican.

What does that tell us? It tells us he’s a Democrat when the votes make it to the floor, but in terms of actually formulating and designing legislation he’s willing to go out of his way to try and get bills on the floor that both sides can agree with. They don’t always make it to the floor, but he clearly works very hard to try and join up with Republicans when it comes to legislating.

So no, Tom Curry, Obama’s Senate record doesn’t place him as a partisan Democrat. In fact, it shows the opposite. It shows a strong desire to bring the two sides together.

Of course, Curry’s article exposes another big problem with this. What does it mean that he “voted with Democrats”? Here we have our explanation:

CQ didn’t use every single roll call vote to make this determination, instead it analyzed votes where a majority of Democratic senators opposed a majority of Republican senators.

In 2007, Obama voted with his fellow Democrats 97 percent of the time. In 2006, his score was 96, and in 2005, he again netted a 97 percent rating.

Once again, there’s a difference between aiming for unity and betrayal of party. If Obama HAD hopped sides in those cases, it would have meant voting against a majority of Democrats.

Let’s review that for a moment. Congressional Quarterly finds when a majority of Democrats vote against a majority of Republicans and discovers that Obama usually sits with the Republicans. Well, yeah. Is that a surprise. It’s a majority of Democrats in a particularly divisive Senate.

You can read CQ’s article here, and note that they don’t exactly define what “a majority” means. Does it mean when 51% of the Democrats voted against 51% of the Republicans? For that matter, how often did that happen?

In an attempt to replicate the study, I hopped on over to the website for the US Senate and took a look at all of the Roll Call votes for the given Congresses. In 2008, of the 27 roll call votes this year, not even half were good enough to fit these criteria. Of the ones where Obama actually voted, they were either blaringly obviously “you’d better vote with the goddamn Democrats” votes like striking telecom immunity from the FISA bill, or instances where it was Democrats to Republicans 100% to 100% almost to a man such as an attempt to invoke cloture a few times.

Now that’s just 2008. So I hopped back to 2007 and combed through a bit as well (though without the time to go through all 442) and found it to be generally similar. Obama did vote with the party most of the time, but that was also because nearly every time it was majority Dems versus majority Republicans, it was a party-line divisive vote. If he had not voted with the party, it would have meant quite literally turning against roughly 95% of the Democrats.

CQ says the following:

In fact, both of them joined their fellow Democrats in mostly party-line roll calls more often than their own majority leader, Harry Reid of Nevada. In the past year, Clinton voted with her party on 98 percent of the questions that pitted a majority of Democrats against a majority of Republicans, while Obama’s score was 97 percent. Reid sided with his party on only 95 percent of those votes.

Makes it seem like Reid is more willing to work with the Republicans, right? Not quite. Reid, when he did side with the Republicans, nearly solely did it in voting against the invocation of cloture, which is a procedural motion he’d take in order to allow himself to bring the issue up again later.

So CQ took his roll call votes and obfuscated the results to make it sound like Obama is “the same as Clinton”, ignoring the fact that by the same criteria, all of the Republicans are the same and all of the Democrats are the same with few exceptions. When the Senate did split by party, it was down the line. All blue against all red.

Curry’s article conflate’s Obama’s “1,100 votes in the Senate” with the 97% number, making it sound like he votes with Democrats and against Republicans that often. However, this ignores two important points:

  1. Those 1,100 counts all of the times where the vote was NOT split along party lines or was accepted/rejected by a vast majority, which a random sampling showed me was quite frequently.
  2. When the vote was split along party lines, it was nigh universal. It was team against team.

When you dig into the study, all you really find is that when a vote was split down the aisle, Obama voted with the other 50 Democrats instead of against them. It would have been downright bizarre for his percentage to be lower, since it would have meant that a significant amount of the time he would be the one Democrat voting with the GOP.

So at the end of the day, Tim Curry misrepresented a piss-poor study that attempted to make Obama sound like a doctrinaire partisan, when a reasonable way of ascertaining that shows the exact opposite to be true.

Just another day in the “liberal media”, eh?

2 responses to “How to screw up “analysis”, by MSNBC.

  1. Hi! I was surfing and found your blog post… nice! I love your blog. 🙂 Cheers! Sandra. R.

  2. Sign: zdbrw Hello!!! bhmnz and 7065fghkhzwemd and 1620 : I will try to recommend this post to my friends and family, cuz its really helpful.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s