It's not news that the Republican Party's popularity is in the dumper, but a recent Gallup analysis hammers the point home by breaking down Repub ratings into 26 subdivisions of the U.S. population.

Comparing the average party ID rankings in all Gallup survey so far this year (which shows a very impressive 53-39 Dem advantage in party ID, including leaners), with an amalgam of similar surveys taken throughout 2001 (when the Dems held a statistically insignificant 45-44 edge) demonstrates how much ground Dems have gained.

But the really devastating analysis, published yesterday, shows that Repubs lost ground in 25 out of 26 demographic subgroups, by race, region, income, education, marital status, etc. The Repubs didn't gain ground in any of the subgroups, but did hold even among frequent church attenders.

Click through to the original Gallup piece and you can see all this in bar charts, but to summarize: The groups with which Repubs suffered the greatest decline (followed by the number of percentage points of drop, between '01 and '09, in the percentage of each group that identify as Republicans) were:

• College Grads: 10-point drop.
• 18-29-year olds, lower-income, middle-income, Midwesterners, moderates and infrequent church attenders : 9-point drop.
• Liberals, unmarrieds: 8-point drop.
• Upper income: 7-point drop.

The best groups for Republicans over this eight-year period were:

• Frequent church attenders: held even.
• Conservatives, age 65 and older, non-whites: 1-point drop.
• Blacks: two-point drop.
• College non-grads: three-point drop.

It's true that senior citizens typically have higher voter participation rates than younger Americans. But it is especially daunting to a party considering its middle- and long-term future when they are doing much better among the young than the old. Political science research suggests that most Americans form a party identification as young adults, and most never change.

Jeff Jones, who wrote the analysis, told me he chose 2001 simply to compare the first year of the G.W. Bush presidency with the first year of Obama. He said the results would have been generally similar if he had taken any of the next several years, when Repubs were still very competitive in party ID. I wondered whether Setp. 11, 2001 might be a confounding factor, creating a temporary high for Republicans that makes the fall to today's ranking look artificially steep.

Jones said not really. Although Pres. Bush's approval rating went from mediocre to the stratosphere in the last quarter of 2001, the party ID numbers didn't change dramatically. In the first quarter of 2001, Repubs had an average party ID rating of 44 percent. In the fourth quarter: 46. This is a reminder that when, on Sept. 12, 2001, Americans rallied around the president, they mostly didn't rally around his party.

One last note: When Gallup reports party ID among three groups, Dem identifiers, Repub identifiers and self-described independents, the gap narrows. For the 2009 ratings, it's:

• Dems: 36
• Repubs: 27
• Independents: 37.

When independents are pushed to indicate which way they lean, it's:

• Dems: 53.
• Repubs: 39.

Republicans might take comfort from the narrower gap when independents are included separately, but, as they contemplate how they are going to rebuild their party, they must also consider why they are losing among indpendents by a 3-2 ratio. During the earlier 2000's, Jones said, independents tended to break roughly evenly.