What Liberal Revolt? This one

by: Chris Bowers

Tue Jan 05, 2010 at 19:49


Mark Blumenthal (of Pollster.com and Mystery Pollster fame) has a column in National Journal, called "What Liberal Revolt?," arguing that President Obama does not have a problem with his liberal base.  The basis for Mark's argument is that President Obama's approval rating has declined less among self-identified liberal Democrats than among any other partisan-ideological subgroup.

Mark's argument that President Obama's approval rating has declined less among liberal Democrats than among any other partisan-ideological group is correct.  Looking at Gallup's tracking poll data over the past year, here are the declines in President Obama's approval rating among Gallup's partisan-ideological groups:

Obama job approval, comparing current levels to peak levels
Group Peak Approval Date Peak Approval % Current Approval % Decline
Lib Dem Apr 20-May 17 95.5% 87.0% 8.5%
Mod Dem Apr 27-May 24 90.8% 81.8% 9.0%
Con Dem Apr 13-May 10 86.8% 74.5% 12.0%
Pure Ind Apr 20-May 17 55.8% 44.0% 11.8%
L/M Rep Jan19-Fed 15 52.5% 30.8% 21.7%
Con Rep Jan19-Fed 15 27.3% 12.0% 15.3%
Note: I used a four-week average in these numbers to smooth out statistical noise.

However, these numbers do not necessarily prove there is no left-wing revolt against President Obama.  For starters, there remains the open question of how accurately these ideological self-identification labels are in determining who is left-wing in America.  Quite a few self-identified moderates might actually hold more left-wing policy positions than many self-identified liberals.

Additionally, it does not take a particularly close look at Gallup's data to see that President Obama does much better among liberals who self-identify as Democrats than among liberals who do not self-identify as Democrats (referred to as LNDs," liberal-non-Democrats" for the rest of the article):

  1. According to Gallup, President Obama's current approval among all self-identified liberals is 77.5%, compared to 87.0% among self-identified liberal Democrats.

  2. During 2009, Gallup identified 21% of the country as self-identified liberal.

  3. In his National Journal article, Mark Blumenthal writes that Gallup shows 15% of the country as both self-identified liberal and self-identified Democrat. This means that, according to Gallup, about 6% of the country is an LND.

    Therefore...

  4. According to Gallup's data, President Obama's approval among LNDs is only 54%.

    Furthermore...

  5. This is down from 70% in February according to Gallup's data, when this group composed only 4% of the population (in his article, Blumenthal says that self-identified liberal Democrats composed 17% of the adult population back in February).
All of that is a significant drop.  President Obama's approval has dropped more among LNDs during 2009 (16%) than among any other part of the American center-left coalition.  Further, this decline has happened even while the group was growing in size, taking in defectors from the liberal Democratic group.

But wait, there's more!  Consider President Obama's current approval rating among ideological groups compared to his vote percentages among ideological groups in 2008:

How has the Obama coalition changed since 2008?
Group 2008 Obama % Current Approval % Change
Lib 89% 77.5% -11.5%
Mod 60% 60.0% 0.0%
Con 20% 28.8% +8.8%
Since President Obama's election, his support among liberals has dropped, his position among moderates has remained the same, and his support among conservatives has increased.  From this perspective, the only people who have become disillusioned with President Obama are liberals.  Everyone else who opposes him never supported him in the first place.

Further, there is a real chance that the decline in President Obama's approval rating among LNDs will translate into a decline in real votes among LNDs. In 2008, liberals were actually the largest swing voting block.  This makes any belief that the liberal voting block is static or "locked in" for Democrats both naïve and stupid.

The bottom line is that, while President Obama's approval rating remains very high among both liberals and Democrats (and particularly high among liberal Democrats), he has suffered a significant drop among LNDs (liberals who do not self-identify as Democrats).  Further, his drop among this group appears to be the only loss he has suffered from his winning 2008 coalition.  As such, despite this group's small percentage of the population (5-6%), this revolt is quite newsworthy and also represents one of the most significant threats to President Obama's re-election that emerged during 2009.

Chris Bowers :: What Liberal Revolt? This one

Tags: , , (All Tags)
Print Friendly View Send As Email

Revolt? (0.00 / 0)
Tea parties - those are revolting.

Liberals are mildly disaffected.


I am guessing many ldn's are people who (4.00 / 1)
used to self-identify Democrat.  I wish they polled for self-identified progressives. I have become disenchanted with the word liberal since I found out the originals were right wing libertarians and since it became associated with neoliberals.

My blog  

You'd think (4.00 / 2)
This article wouldn't even need to be written after 2000.  

You'd think (4.00 / 2)
a Democratic President who campaigned on "Change" would have chosen Main Street over Wall Street.

Life's funny that way.


[ Parent ]
but after 2000, we established (0.00 / 0)
that any lnd is a traitor that hates america, and would rather have a republican destroy the country than make a compromise on anything.  Ergo, their opinions are worthless and irrelevant.

or something.


[ Parent ]
probably not news to anyone (0.00 / 0)
ok let me get this straight, 16% out of a total of 5% don't like the prez anymore....

Looks like you're trying to fight the manifest data (0.00 / 0)
Hey, I think that first poll is pretty decisive in making the opposite point that you're trying to make.

For starters, there remains the open question of how accurately these ideological self-identification labels are in determining who is left-wing in America.  Quite a few self-identified moderates might actually hold more left-wing policy positions than many self-identified liberals.

In the absence of any evidence that filling in this "open question" data would change the results in favor of your theory, you should expect that it doesn't. Is there any reason why you're confident that it would, that's more than a hunch based on anecdotes?

Additionally, it does not take a particularly close look at Gallup's data to see that President Obama does much better among liberals who self-identify as Democrats than among liberals who do not self-identify as Democrats....

Well duh, I think you will in general find Obama to be more popular among Democrats than non-Democrats in any sub-group, him being the obvious leader of the Democratic party. How should this obvious fact make your case?

But really, here's what seems most irresponsible: Your comparison of voting proportions [apples] to current approval numbers [oranges]. You know very well the well-studied reasons why voting and approval generally diverge. The "effect" you "discover" can be accounted for entirely by the fact that people are more likely to approve of the leaders of both parties than they are to disapprove of both - but these people still vote for their own.

In general, the argument in this post is quite bad.


[ Parent ]
You're ignoring the actual kicker in the data (4.00 / 2)
when you group by lib-mod-con, as Chris did in the bottom table, you see a decline ONLY amongst liberals.  That is important.  The rest of the data in the article is mostly anecdotal language building up toward explaining THAT upshot.

[ Parent ]
Small sample subsize? (0.00 / 0)
I'm as annoyed with Obama as any other liberal, but from a statistical point of view, I'd hesitate to draw such strong conclusions from what are very small sample sizes. A 16% swing from a 5% subsample could be nothing more than statistical noise.  

Core customers and trendsetters (4.00 / 2)
This article about how Whole Foods' CEO John Mackey erred in his healthcare editorializing by offending key trendsetters in his customer base seems relevant.

Mackey ticked off the people whose steady pressure over many years created his customer base in the first place. They aren't a majority, they aren't enough of a percentage to cause major losses in sales. But their bad opinion is enough to cause stock prices to drop, enough to undermine future market share expansion potential.

The people who saw the possibility of all food markets being in some respect like Whole Foods before it existed will either keep this one going strong or make it possible for the next Whole Foods to eventually overtake it. Like water on a stone.


Old polling methods may not apply .. (0.00 / 0)
..nor old labels.

There's a whole lot of very shaky gray matter there not well represented or understood.   With the Democrat's abject gutlessness combined with the Rights insanity we're entering some very strange times filled with the paranoid politics not unlike those of the 70's, out of which sprang President Richard Nixon.

In an interesting book on the 70's by Francis Wheen, "Strange Days Indeed", the mindset and direction of everyone after the catastrophes of the 60's could never have been predicted.
But it's Wheens earlier book that we should all read as it may pave the way to a better understanding of what's to come:
"How Mumbo-Jumbo Conquered the World."

http://www.guardian.co.uk/book...


Nationalism is not the same thing as terrorism, and an adversary is not the same thing as an enemy.


USER MENU

Open Left Campaigns

SEARCH

   

Advanced Search

QUICK HITS
STATE BLOGS
Powered by: SoapBlox