Are Exit Polls the only basis for the numbers the Pollsters put out about

how many whites, blacks, latinos and working class vs. college educated vote for a candidate? If it's only based on "exit polls" can we really believe them?

What other basis would pollsters have for knowing the exact demographic of voters?

Anyone know? Is it just Exit Polling?

Thanks if you have any info...and not snark or silliness.

Analysis like this for example:

----------------------------------

Obama Victorious, Clinton On The Ropes

Thomas B. Edsall
The Huffington Post

The demographic patterns on Tuesday suggest an intensification of racially polarized voting.

In the March 4 Texas primary, Clinton won whites (46 percent of the total), by 55-44; she won Latinos (32 percent of the total) by 66-32, while Obama carried blacks by 84-16.

In Ohio, also on March 4, white Democrats (76 percent of the primary turnout) backed Hillary 64-34, while blacks (18 percent of primary voters) supported Obama 87-13.

Seven weeks later, in the April 22 Pennsylvania primary, white Democrats (80 percent of primary turnout) voted 63-37 for Clinton while blacks (15 percent of the total) votes 90-10 for Obama.

"It looks like a big win for Obama in North Carolina and a narrow win for Clinton in Indiana," said Emory political scientist Alan Abramowitz as the first exit polls were released. "That's not good for Hillary. She needed a breakthrough -- a big win in Indiana and a win or a narrow loss in North Carolina. It looks like she's not going to get either. Obama will add to his delegate lead."

It is unlikely, however, that Clinton will give up at this stage.

"I can tell you right now . The battle goes on, a fight to the death, one delegate at a time, never say die, millions of Democratic voters yet to be heard, white working class people will vote for McCain instead of 'him,' blah, blah, blah," declared Lawrence F. O'Brien, Democratic lobbyist, donor, and the son and namesake of the Democratic National Committee chair in the late 1960s and early 1970s.

Open Left's Chris Bowers wrote: "Given that Obama was already ahead... and that there are now very few states remaining, that is a very good night for him....it is just what he needed to help turn around the media narrative."

Clinton has demonstrated an extraordinary will to win, and a refusal to quit when she was losing primary after caucus after primary. Her campaign is now arguably at the stage where it is dependent on miracles -- like the surfacing of a new, and worse, Jeremiah Wright controversy or a Chicago scandal implicating Obama.

Norman Ornstein, of the American Enterprise Institute, contended, in effect, that it's all over but the shouting:

"I have long viewed this in a simple way: two things matter, delegates and popular votes. If Obama wins both, he cannot be denied the nomination. If these numbers hold up, he will erase her gains in Pennsylvania and have a near-insurmountable popular vote lead. That will do it, and I expect a stream of superdelegates to move to him in the coming week-plus."

While it was clearly Obama's night, there were some small glimmers of hope for Clinton in the exit poll data.

In Indiana, she very narrowly beat Obama (51-49) among men, held her own (56-44) with working and lower middle class voters without college degrees, and won among the one in five white, self-described "independents" -- often an Obama constituency -- by a slim, 52-48 margin.

In addition, the 67 percent of Indiana voters primarily concerned with the economy now completely eclipse the 18 percent who give top priory to the Iraq war. Hillary wins the economy voters by 53-47, while Obama carries the Iraq-priority voters by 54-46.

Conversely, the significance of Clinton's victory in Indiana was undermined by indications that a statistically significant number of Republicans, perhaps as many as 7 percent of all the votes cast, were following the suggestion of conservative talk show host Rush Limbaugh to cast ballots for her in the Democratic primary.

Further elaboration can be found in the analysis of my colleague Sam Stein. Just over one in ten Indiana Democratic primary voters was a Republican, a constituency that has backed Obama on other contests, but in Indiana Clinton won them 53-47, possibly as a result of Limbaugh's exhortations.
http://www.huffingtonpost.com/2008/05/07/obama-victorious-clinton_n_100521.html

so all the talking head blather is pretty much blather during those awful analyses

I guess I just don't accept that only "uneducated white women vote for Hillary" and mostly "African-American's and urban intellectuals" vote for Obama. How would pollsters know that right after an election unless it's based on "exit polls" and could those results be skewed by questions asked and the sample?