Funny We Hadn't Heard about This

[tablesaw]

I saw this on Alas! A Blog:

So yesterday I was listening to Patt Morrison on KPCC, and she was interviewing various people about prop 8. One of the topics that came up was, of course, support for prop 8 in the black community. But before she began her interview with Jasmyne Cannick, a black lesbian journalist who wrote an op/ed for the LA Times, she stated that "lack of black support wasn't as substantial as initially reported - just over 50%, very much in line with much of the rest of the population."

—Julie (formerly The Girl Detective), "Hold the *&%$ing Phone"

I couldn't listen to the show, but it did send me on a new hunt for more exit polling, which led me to The Leavey Center for the Study of Los Angeles, which released the results of its exit polling in Los Angeles (PDF file).

Here's what it says on Prop 8:

How did you vote on the following: Prop. 8, Eliminates Right of Same-Sex Couples to Marry. Initiative Constitutional Amendment?

N = 2686	All City (100%)	Region		Ethnicity
		Non-valley (60%)	Valley (40%)	White (55%)	Hispanic (24%)	Black (15%)	Asian (6%)
Yes	30%	31%	27%	19%	48%	45%	34%
No	57%	54%	63%	73%	42%	41%	57%
Didn't vote	2%	2%	1%	1%	3%	2%	0%
No response	12%	13%	9%	6%	7%	12%	9%
Total	101%	100%	100%	99%	100%	100%	100%
© 2008 Leavey Center for the Study of Los Angeles Presidential National Election Exit Poll in the City of Los Angeles. November 6, 2008. Margin of error is ± 1.89% for All City category.

Gee, those numbers look more reasonable don't they? And they sampled a larger number of African-American voters. And they sampled them in one of the cities where African-American voters are more widely represented in the state.

The report said this about its methodology:

This exit poll was part of LCSLA's effort to implement new sampling methodology called the "racially stratified homogeneous precinct approach." This new method addresses limitations in standard exit poll sampling that typically has not provided accurate sampling of ethnic groups in urban settings.

I didn't see specific explanation of what that is, but it does indicate that this exit poll was designed so that it would more accurately represent different groups' voting patterns.

Will this bit of data start to spread? Well this report came out last week, so we've got some work to do now that the truth is getting its boots on.

SatNYTX: 9:30 (wow!); SunNYTX: 12:45 (over 3x faster than Lisa); MonNYTX: 3:15.