A Polling Free-Fall Among Blacks (just 2% support Bush)

[freedomburns]

Talk about a mandate!

http://www.washingtonpost.com/wp-dyn/content/blog/2005/10/13/BL2005101300885.html?nav=rss_nation/special


A Polling Free-Fall Among Blacks
By Dan Froomkin
Special to washingtonpost.com
Thursday, October 13, 2005; 12:24 PM

In what may turn out to be one of the biggest free-falls in the history of presidential polling, President Bush's job-approval rating among African Americans has dropped to 2 percent, according to a new NBC/Wall Street Journal poll. The drop among blacks drove Bush's overall job approval ratings to an all-time low of 39 percent in this poll. By comparison, 45 percent of whites and 36 percent of Hispanics approve of the job Bush is doing.

A few months after the 9/11 terrorist attacks, the NBC/Wall Street Journal poll found Bush's approval rating among blacks at 51 percent. As recently as six months ago, it was at 19 percent. But Bush's bungled response to Hurricane Katrina -- seen by many blacks as evidence that he didn't care about them (see my September 13 column ) -- may have brought support for the president in the African American community down to nearly negligible levels.

Tim Russert called attention to this startling statistic on the NBC Nightly News with Brian Williams yesterday: "Brian, listen to this," he said. "Only 2 percent -- 2 percent! -- of African-Americans approve of George Bush's handling of the presidency -- the lowest we have ever seen in that particular measure." So this morning, I called Democratic pollster Peter D. Hart, who conducted the survey with Republican pollster Bill McInturff, to get a better sense of the significance of the results. "African Americans were not supporters, but I don't think that they outright detested him -- until now," Hart said. "The actions in and around Katrina persuaded African Americans that this was a president who was totally insensitive to their concerns and their needs." Hart said he has never seen such a dramatic drop in presidential approval ratings, within any subgroup.

This latest poll included 807 people nationwide, and only 89 blacks. As a result, there is a considerable margin or error -- and the findings should not be considered definitive until or unless they are validated by other polls. David Bositis, a senior political analyst at the Joint Center for Political and Economic Studies, which tracks African American public opinion, told me this morning that it's clear that Bush's job approval among blacks "has taken a hit from both the ongoing things in Iraq and what happened with Katrina." But down to 2 percent? "I doubt that it's actually 2," he said. "But would I be surprised if it's 10 or 12? No." And 10, he said, is typically "about as low as you can go" when it comes to approval ratings.

more.......

http://www.washingtonpost.com/wp-dyn/content/blog/2005/10/13/BL2005101300885.html?nav=rss_nation/special

Sure, it's a small sample of black voters.  What's that mean?  It means the percentage could be smaller.  Bush could really have 0% support among blacks right now.  The internal statistics on the poll are here: http://online.wsj.com/public/us

fb

[Jake]

Do coloreds matter anyway?

[○∙◄☻¥tπ[╪AV┼cVê└]

Do coloreds matter anyway?

Are you trying to get this to drop to 1%?

[opebo]

Do coloreds matter anyway?

Good point.  But I suppose if the GOP gets 2% of the black vote in 2006 or 2008 instead of whatever it got with the fake war on terrorism/patriotism nonsense in 2004, it will make a difference in a couple of states.

[TheresNoMoney]

Bush got something like 11% or 12% of the black vote in 2004.

I will gladly let the Republicans take their 2% in 2008.

[freedomburns]

Do bigots matter anyway?  

[Blue Rectangle]

Bush got something like 11% or 12% of the black vote in 2004.

I will gladly let the Republicans take their 2% in 2008.

From the same Washington Post article:

You must be logged in to read this quote.

So, basically this story is meaningless.

[○∙◄☻¥tπ[╪AV┼cVê└]

MOE is smaller on 2% out of 89 than 12% out of 135.

[Blue Rectangle]

MOE is smaller on 2% out of 89 than 12% out of 135.

What are the MOEs?

[○∙◄☻¥tπ[╪AV┼cVê└]

MOE is smaller on 2% out of 89 than 12% out of 135.

What are the MOEs?

2% = 2 out of 89 gives 2*sqrt(2)= 3%
12% = 16 out of 135 gives 2*sqrt(16)=6%

[Alcon]

Very few polls are large enough that their black sample size is significant.  Besides, unless regionality is provided, it is fundamentally useless.

[Fmr. Gov. NickG]

MOE is smaller on 2% out of 89 than 12% out of 135.

What are the MOEs?

2% = 2 out of 89 gives 2*sqrt(2)= 3%
12% = 16 out of 135 gives 2*sqrt(16)=6%

Margins of error are based on the total subsample sample size (89 and 135).  So the margin or error is greater in a sample of 89 than a sample of 135, no matter how many people answer one way or the other.

[Fmr. Gov. NickG]

BTW, the margin of error for a sample of 135 is about 9%; the MoE for a sample of 89 is about 11%.

[A18]

Okay, so basically we have no clue what black support for Bush is.

[Fmr. Gov. NickG]

Okay, so basically we have no clue what black support for Bush is.

Well, we know it's pretty small.

[Blue Rectangle]

BTW, the margin of error for a sample of 135 is about 9%; the MoE for a sample of 89 is about 11%.

If the results were near 50%.  jfern used an approximation for very low or very high percentages.  I got 4% and 7% for 99% confidence.

[A18]

Point is, we know nothing new from these polls.

[Alcon]

Point is, we know nothing new from these polls.

Yes, and a 2-11% support would hardly be a "free-fall."  I believe the only places Kerry received less than 80 percent of the black vote according to exit polls were Oklahoma and Washington state, and he received 88% nationally.

I know the state exit polls shouldn't be trusted, but the national had a solid enough sample.

[J. J.]

BTW, the margin of error for a sample of 135 is about 9%; the MoE for a sample of 89 is about 11%.

Not a great MOE either way.

A drop though could make a difference between PA, for example being in play and PA being a relatively safe state.

[A18]

Well, it depends where the drop takes place. It obviously matters more in Pennsylvania than, say, Utah.

[The Vorlon]

BTW, the margin of error for a sample of 135 is about 9%; the MoE for a sample of 89 is about 11%.

Actually...

The typically quoted margin of error is based upon the two choices (ie approve and disapprove in this case) having fairly equal probabilities (ie each fairly close to 50%)

When you deal with very small versus large probabilities (ie say 90%/10%) your accuracy actually improves.

For example if we assume the "true" support for Bush among African Americans is (+/-) say 90% disapprove, 10% approve, then a sample size of 135 actually has a margin of error right around 3% or so.

(the binomial expansion aproximation of the normal curve is .1*.9=> .09 not .5*.5=>.25 you get at a 50/50 probability)

To think of this intuitively, if you want to know the true probability of flipping a coin (almost 50/50) you have to get a really big sample to determine that it actually is 50/50. - Even with a sample of 1000 about 1/3 of your samples will be out by 3% (ie 515 to 485 or more)

By contrast, You need to flip very few coins to determine that coins usually land on one of their faces, and very rarely stand up on their edge after you flip them...

As a general rule; subsamples this small are always to be taken with a grain of salt... a subsample of say 100 Blacks just is not big enough to be balanced in terms of income, geography, education, age, gender, etc...

It will be very interesting to track Bush's African American support post Katrina, but given all the uncertainties of small sample sizes associated with this portion of the population, we are going to have to see a lot of polls which taken collectively might tell us something.

[A18]

Why is Vorlon red?

[freedomburns]

Why is Vorlon red?

Vorlon has seen the light and swiched camps.  He will soon be changing his name to "The Shadows".

[MODU]

Again, we're talking about polls, which is nothing more than a popularity contest.  Most of the tough/right decisions tend to be the most unpopular ones.  Not to imply that Bush makes all the right decisions, but you get the point.

[dazzleman]

Again, we're talking about polls, which is nothing more than a popularity contest.  Most of the tough/right decisions tend to be the most unpopular ones.  Not to imply that Bush makes all the right decisions, but you get the point.

Presidents who make tough decisions will always go through periods of unpopularity.

When Clinton was trying to actually do things in his first two years, he was pretty unpopular.  When he did nothing in his last 6 years, he was very popular.

The public is like children.  Children love parents who let them eat sweets before dinner, don't give them a set bedtime, and don't make them go to school.  Then, when they grow up unprepared to function in the world, they blame their parents for all the things they loved as children.

Presidents who put their own popularity first do not prepare the public to deal with unpleasant issues, and pass looming problems that maybe could have been contained on to their successors.  That is exactly what Clinton did, and what Bush is attempting to avoid doing.  I don't claim that everything he's done is right, but he has already taken on far more contentious issues in 4 years than Clinton did in 8 years.

As far as Bush's popularity among blacks is concerned, who cares?  The vast majority of blacks wouldn't like him no matter what he did, so there isn't much point in wasting a lot of time fretting over their opinions.