Thank you, Daily Kos, for commissioning a daily tracking poll and providing detailed breakdowns of the internals. But please, please, please don't repeat the mistakes of other pollsters in your breakdowns of the racial backgrounds of the people polled.
To be specific, please stop burying Asian American voters in an undefined "Other" category.
Asian are just behind Latinos as the second fastest growing segment of the US population -- and are increasingly important in key states like Virginia (4.8 percent Asian) and Nevada (8.2 percent Asian). Asian Americans also vote disproportionately Democratic -- even overwhelmingly so, as AALDEF's 2006 exit polls have shown.
Relegating Asians to an anonymous "Other" category denies us of potentially valuable information that could be used to plan, target and inspire outreach to an important and growing segment of Democratic voters.
Furthermore, the first day of the Daily Kos poll's internals reveals a significant undersampling of Asian American voters.
The Daily Kos poll's internals note that 73 percent of the respondents are white, 13 percent are African American, 13 percent are Latino, and 1 percent are "Other/Ref". (I'm assuming "Ref" means "refused to answer.")
But according to the US Census, in 2004, 79.2 percent of voters were non-Hispanic whites, 11.1 percent were Black, 6 percent were Hispanic (any race), and 2.2 percent were Asian (see page 10 of this pdf). And the 2004 CNN exit polls showed an electorate that was 77 percent white, 11 percent African American, 8 percent Latino, 2 percent Asian, and 2 percent Other.
Of course, if the DKos polling crew comes through with the "Asian" category, we'll have to argue about how valuable the data is, since it's coming from such a small sample. But at least the fact that we'll be having that argument will nail home the fact that Asian American voters are a critical part of the progressive coalition and could hold the key to an Obama victory in a number of critical states.