- Jul 28, 2006
- 18,251
- 8
- 0
Since we talk about this so much I thought it wise to share this article with the group.
I think the guy is spot on.
1. Like it or not Obama earned that rating.
2. The rating is not much more than a tool used to compare on Senator to another.
link
I think the guy is spot on.
1. Like it or not Obama earned that rating.
2. The rating is not much more than a tool used to compare on Senator to another.
link
One of the habits I've developed in recent years is to go to Google once or twice a day and type in "National Journal." As editor of National Journal, I figure it's part of my job to stay abreast of the mentions the magazine receives in the press and the blogosphere. The exercise can be rewarding (when we break a big story that gets a lot of attention), humbling (when we break a story that gets very little attention), and -- this year, in particular -- exasperating.
That's because of our vote ratings. Every year since 1981, National Journal has ranked members of Congress on a liberal-conservative scale. Earlier this year, NJ ranked Barack Obama's voting record as the most liberal in the Senate in 2007, a characterization that, not surprisingly, has generated much coverage -- and more than a little criticism.
Unfortunately, much of the criticism has been uninformed. There have been exceptions, such as Obama's own response to the rating and the occasional well-researched article that raises legitimate questions. On the whole, however, the commentary has been pretty superficial.
What follows is an effort to respond to some of the objections.
First, a bit of background.
Toward the end of every year, a panel of National Journal editors and reporters sifts through the hundreds of roll-call votes taken that year and selects ones that we think are useful in identifying ideological differences between members of Congress. For 2007, we ended up with 99 votes out of the 442 votes cast in the Senate. We sent the votes to the Brookings Institution, which is under contract to National Journal to compute the vote ratings based on a system designed by William Schneider, a CNN political analyst and a contributing editor to the magazine. After Brookings delivers the ratings, we run them in our magazine and post them on our Web site.
To dispense with the silliest criticisms first: National Journal did not rig the ratings so that Obama would be ranked as the most liberal senator. NJ is not a right-wing publication out to get Obama. And we're not reveling in the attention that the ratings have generated.
To the contrary. We don't have any idea when we send the votes to Brookings how an individual member of Congress will be ranked. We make every effort to be nonpartisan and nonideological in our news coverage. (We do have columnists who take liberal or conservative positions.) And, at least as far as I'm concerned, the attention has been a headache. My first reaction when I learned that Obama had scored the highest liberal rating was to mutter a profanity. I had an inkling of what was in store.
Other criticisms of the rating system deserve more of a response.
? The ratings are flawed because every vote isn't counted.
We don't count every vote, because many have little to do with ideology. What would be the liberal or conservative position, for example, on Senate roll-call vote 236 on a measure to provide for the care and management of wounded warriors? The measure passed 94-0.
? Obama wouldn't have been ranked as the No. 1 liberal if he hadn't missed so many votes.
Perhaps that's correct, but how would anyone know for sure? To qualify for a vote rating, a member of Congress needs to participate in half of the selected votes. Obama cast votes in 66 of the 99 roll calls that formed the basis for the Senate ratings. On those 66 votes, he took what we labeled the liberal position 65 times. Whether Obama would have taken the conservative position on some of the votes he missed is a matter of conjecture.
? Some senators took the liberal position more often than Obama. So he couldn't have been the most liberal senator.
Obama's No. 1 ranking is akin to being declared the major-league batting champion. The honor goes to the player with the highest batting average, regardless of whether he has the most hits. In Obama's case, voting the liberal position 65 out of 66 times earned him the title, as opposed to a senator who might have voted the liberal position 80 times out of 90.
? Anyone who follows the Senate knows that Obama isn't as liberal as, say, Sens. Bernie Sanders of Vermont, Barbara Boxer of California, and Russell Feingold of Wisconsin.
It's true that those senators are generally perceived as being more liberal than Obama. But our ratings are based on actual votes, not perceptions; and in 2007, Sanders, Boxer, and Feingold cast a slightly lower percentage of liberal votes than Obama did. Feingold, for instance, was one of only two Democrats who voted to uphold President Bush's veto of the Water Resources Development Act. Boxer voted with 37 Republicans and only six other Democrats on a proposal regarding Real ID requirements for driver's licenses. Sanders voted with 37 Republicans and only 15 Democrats against limiting debate on the immigration reform bill. Just because someone is perceived as a liberal doesn't mean he or she votes the liberal position every time.
? Some of the votes selected for the vote ratings don't lend themselves to being classified as liberal or conservative.
This, in my view, is the most cogent criticism of the vote ratings, and it is the argument that Obama himself made when asked about his ranking.
In February, Obama responded to a question from Politico Editor-in-Chief John F. Harris about the vote ratings by saying, "An example of why I was rated the most liberal was because I wanted an Office of Public Integrity that stood outside of the Senate, and outside of Congress, to make sure that you've got an impartial eye on ethics problems inside of Congress. Now, I didn't know that it was a liberal, or Democratic, issue. I thought that was a good-government issue that a lot of Republicans would like to see."
Few analysts would challenge most of the votes selected for the ratings (such as those on abortion rights, the minimum wage, and the estate tax). On a handful of votes, however, reasonable people can disagree about their inclusion. The vote that Obama cited was one of them.
My rationale for selecting the vote was that efforts to reform Congress have long been a hallmark of liberals, from Sen. Robert LaFollette of Wisconsin, to the Watergate class of 1974, to Sen. Paul Wellstone of Minnesota. The vote on the Office of Public Integrity, in my view, was in that tradition.
It should be noted, however, that just because the National Journal panel thought the vote was worthy of inclusion doesn't mean that it was automatically counted. Under our system, any vote that NJ editors and reporters select can be knocked out when Brookings crunches the numbers.
How so?
Clearly, each chamber has liberal and conservative camps. The most-conservative members vote together most of the time, and the same is true of the most liberal members. Our rating system tags the liberal and conservative camps by identifying the members who most frequently voted with one another. Each vote is then weighted based on how tightly each camp stuck together. If the computer analysis finds that the liberal and conservative camps are indistinguishable on a vote, the vote is kicked out.
In the case of the vote on the Office of Public Integrity, the computer analysis (called a principal-component analysis) found enough correlation to use the vote.
One last observation: In the end, the debate over whether Obama was the most liberal senator last year doesn't strike me as particularly useful. If Obama had voted differently a couple of times, or if we had added or subtracted a couple of other votes in our ratings, he might not have been ranked as the most liberal senator. Perhaps he would have ended up as the fourth-most-liberal senator. Or maybe the sixth-most-liberal senator. (In his first two years in the Senate he was ranked as the 16th- and 10th-most-liberal senator.) But by virtually any measure, he still would have had a solidly liberal voting record -- one that would give some people a reason to support him and others a reason to oppose him.
As I said in an editor's note when the 2007 vote ratings were released, they should be viewed as a tool in assessing a member of Congress, but not the only tool. Other vote ratings should also be taken into consideration, as should attributes that no rating system can measure, such as character, judgment, effectiveness, leadership, and experience.