Probably the best thing David Frum’s Frum Forum (say it three times fast) does is occasional podcast interviews with Jay Cost. Cost did one of them earlier this week and it’s gold. There’s so much to love about Cost, from his command of congressional politics minutia to his serious funny (“what really bakes my clams is . . .”). But what I like most is that Cost is (1) Open about having his own policy preferences while (2) Firmly believing that his policy preferences are irrelevant to how the political world works.

One of the biggest problems with punditry (maybe the biggest problem?) is how writers align their preferences with their situational analysis: Pundit A favors Legislation X and therefore believes that Legislation X will be wildly popular and beneficial to Pundit A’s party. My favorite example of this is the Sarah Palin Rorschach Test: Whether a pundit believes that she is a serious contender with a plausible path to the Republican nomination and/or White House depends entirely on whether said pundit likes or dislikes Sarah Palin.

Anyway, Cost will have none of that. He’s just fantastic.

Incidentally, he touches briefly on something I wrote about this week: The disparity in job approval rating numbers between black respondents and the rest of the country. In case you’re interested, here’s the nub:

Obama’s black job approval numbers are more than double his overall [job approval] numbers. What that means is that the level of support the president receives from this group moves the overall number more than you might imagine. When you do the math, accounting for percentages of population (roughly: 75 percent white, 12 percent black, and 13 percent Hispanic/other), you find that today the black vote moves the overall number significantly. Using Gallup’s data, blacks push Obama’s overall number up by about 5 points; using Rasmussen’s by roughly 7 points.

Now all core supporters move the overall number of their candidate upwards; that’s why they’re called a base. In a presidential election, this trend has few ramifications. The presidency is a nationally elected office, and nationwide approval indices are a good measurement of the prospect of reelection. But this skewing of the president’s job approval number creates complications for congressional candidates. While about 12 percent of Americans are black, relatively few congressional districts have an average demographic make up. Because of gerrymandering, mandated majority-minority districting, and simple geographic diversity, blacks tend to be concentrated in certain congressional districts. There are 31 districts with a black population over 40 percent. Only 132 districts are above the national average in terms of black population—leaving 303 districts below the national average.

This racial concentration creates a great many districts which are significantly less black than the nation as a whole. For instance, 62 districts are less than 2 percent black; 107 districts are less than 3 percent black; 177 districts are less than 5 percent black. The median congressional district has a black population of only 6.41 percent. 

This uneven dispersal magnifies the disparity of approval between Obama’s base and the rest of the country. If relatively few congressional districts look like America, then in most congressional districts Obama’s job approval is likely to be lower—anywhere from 2 to 7 points lower—than the national average. (Conversely, in a smaller number of districts it is likely to be much, much higher.) . . .

It’s not hard to see why this phenomenon might concern the folks running Democratic campaigns. Charlie Cook has 23 Democratic-held seats currently rated as toss-ups, and this sample looks a lot like Congress as a whole. Only six of the 23 have black populations above the national average and in five of these districts, as you might expect, the black population is over 20 percent. But of the 23 districts, the median black population is only 5.67 percent. Eleven of these seats have a black population under 5 percent. In seven of them the black population is under 2 percent.

Click through for greater detail and some historical perspective.

Update: Cost has a new post up filled with great nuggets, like this:

Witness, for instance, the number of members who are defecting on minor procedural matters. For instance, seventeen brave House Democrats voted with the Republicans on the highly controversial resolution yesterday to adjourn the House of Representatives! That includes 10 Democrats who just voted for ObamaCare but who were courageous enough to defy the Speaker’s demand to send members home for Easter vacation: Chris Carney, Joe Donnelly, Brad Ellsworth, Jim Himes, Suzanne Kosmas, Harry Mitchell, Scott Murphy, Tom Perriello, Mark Schauer, and Joe Sestak.

Kidding aside, there is no other reason for such a vote than to lower the percentage of agreement with Speaker Pelosi.