Articles and Analysis


Exit Poll Data: Education and Race

Topics: 2008 , Barack Obama , Exit Polls , Hillary Clinton , Jon Cohen

The Washington Post's Jon Cohen has posted some extremely useful data from the exit polls to his Behind the Numbers blog. He ran the Clinton-Obama vote by education among white voters and found evidence of a large "education gap:"

In each of the states where the Post subscribed to exit polls (and voters were asked about their level of education), Clinton did better among non-college than college-educated white voters. She also outpaced Obama among non-college whites in all 14 of these states, but beat him by more than a single percentage point among college graduates in only five.

This data helps shed some light on the subject of speculation earlier this week, whether Obama is "finally cracking the code of the working class white voter" as one observer put it. Our initial look at the exit polling on this issue was inconclusive, because the official exit poll tabulations show the results by education (and income) among all voters. Since the biggest differences between Clinton and Obama have been by race and ethnicity, the share of African American or Latino voters in each state determines whether they do better or worse among the less-well educated voters in that state.

Cohen's tabulations control for race, showing the percentage by education among white voters in each state, thus allowing for better comparisons across states:

02-15 Post Exit Poll Data

Obama's share of non-college whites in Virginia was, as many assumed, higher than in any other state except Illinois, although his performance among this subgroup has been relatively consistent elsewhere. Obama's percentage of non-college whites in Maryland was similar to most of the other states. Also, as some have speculated elsewhere, his percentage of non-college whites was lowest in three Southern states: Florida, South Carolina and Tennessee.

Somewhat surprising -- to me at least -- is the much larger variation across states among college educated white voters. Obama had large double digit leads among college educated white voters in Virginia, Missouri and Illinois but trailed by double digits among college whites in New York, New Jersey and Florida.

Some of these differences are clearly related to the home state advantages (Illinois, New York and possibly New Jersey). Others may have to do with the relative expenditure of resources (candidate time, television advertising and field organizing) by Obama and Clinton in each state. Do our readers have other theories?



I dont have the data easily available, but is there any discernable pattern in white support by education across either closed or open primaries? Does Obama generally do better among white voters, either of low or high education, in open primaries as he may draw more white independent voters?


Richard Pollara:

I think your post and Margie Omerio's earlier today cut into the media perception that there has been a movement from Clinton's core groups to Obama. It has seemed odd to me that a large demographic shift would occur in the 7 days after Super Tuesday. I think this race is likely to end up as a tie (witness the national polls). Obama does well when his core groups vote and Clinton succeeds when hers cast their ballots. Looking at the numbers and accounting for all of the different variables it seems that the support base for each candidate is much more stable than previously thought. If this is true we are liable to see an evenly divided race after April 22. I think Democrats need to start thinking NOW about how they will break the tie without destroying the eventual nominee.



Reading this I had the exact same question as erik.

Specifically, is there any correlation between Clinton or Obama doing better when it is an open primary? Is there any correlation when voters are allowed to registered on the day of the vote?

If there is a correlation, this may have some impact on the expected results of Wisconsin.



I think this analysis is missing a larger picture.

White Democrats have consistently gone for Clinton in non-caucus states. Obama has been closing the gap slightly in that category, but most of his perceived increase in drawing the White vote is from winning over White Independents. Closed primaries, like PA, don't bode well for him for this reason.




an update to my own question, of the primaries listed, only four were closed - Florida, Arizona, New York, and Maryland

So if you did a very unscientific and statistically questionable analysis of
these four closed primaries compared to the rest:

Clintons avg. advantage among non-college whites is 33% among the closed states and 27% among open states - though this includes Florida which is really an outlier. If you remove Florida, then the comparison is really 30% vs. 27% - not really any difference

Among college-educated whites, the avg. clinton advantage is 9pts among closed states (6pts if you exclude Florida) and -5.2pts among open primary states.

So, "possibly" - open primaries make a difference for Obama's support among white college-educated voters...but since New York was Clinton's homestate and Florida's election is not truly representative compared to the others...this analysis is shaky at best. More data, as always, is needed.



Daniel T:

I will go with the comment I just posted in the other thread: this is making a mountain out of a molehill. Even once we factor in race, we see that Obama still attracts non-college voters. Only in four states does he poll less than 30%. In some states Hillary does *marginally* better and in others Obama does *marginally* better. But it simply a twisting of the data to say in an absolute sense that Obama cannot attract non-college and Hillary can. We are simply talking matters of minor degree and without doing a complete correlation, impossible to know just how much education matters in the bigger picture.

BTW, why is Florida in these results? It was not a valid poll. No one really competed and the votes don't count. I think it unwise to include Florida in any results.


Daniel T:

"In Post-ABC national polling, Clinton consistently scored better among voters without college degrees than among those with more education, and the pattern has held firm in primaries across the country. In fact, education has been a *key divider* among white voters in a contest marked by an evident racial divide."

That statement is simply not supported by the data. He can't say for certain that education is a "key divider" w/o control for all other factors, which he did not do. This is either bad reporting, ignorance, or deliberate obsufcation.

I wouldn't trust anything this guy says.



Is Hawaii going to be polled by Edison Mitofsky?


Linda S.:

I think Daniel T. is right about not trusting anything this guy says (i.e. Cohen). The spin here appears to be that less educated whites are all racist. I think that conclusions such as his based on exit polls should be thrown in the trash pile of what the MSM wants us all to believe.



Conservative white voters go for Obama because he is much more conservative in tone: he is the only Democrat I know of who believes that Pres. Bush's motives were good in destroying our civil liberties. Obama both wanted to vote for John Roberts and defended those who did. Obama backed Lieberman strongly in the primary thereby dealing a much greater blow to the antiwar movement than any vote by Clinton.



To take up a point from an earlier discussion, it's clear that Obama did well among whites in VA both because they are mostly educated and because he did better among those who are not.

Looking forward to Ohio, what are the expected turnouts for the two groups? One especially interesting case, on its own and in relation to Ohio, is Missouri. Obama narrowly won MO despite a very high % of less-educated white people (48%), and without doing particularly well among that large group. Instead, he dominated among college whites, an unusually large education gap. I wonder what created this.

MO is particularly relevant to the OH picture because they are both Midwestern states that border the South, both have few Latinos and about 10% African-Americans, both have mixed rural-urban populations.



Mr. Blumenthal,

First, I want to thank you for all of the analysis you are doing on this site. Its great!

I am curious what these tables would look like if we consider the african-american vote. That is, is there any difference between college-educated and non-college african-american voters?

As for the white vote, I find this table extremely telling. We see that Senator Clinton maintains an advantage among non-college voters everywhere, even in Illinois! Senator Obama's support from college-educated white voters is much less geographically consistent; he did not carry them in several northeastern states (although I am sure he carried them in CT even though it is not listed) nor in some states in the south.

Senator Clinton's average margin of victory in this set for non-college voters is 28.5%; Senator Obama's average margin for educated voters is 1%. If we exclude Florida, these averages change to 27.5% and 2.5%, respectively, with Senator Clinton's average margin still much greater in "her" demographic. It would be interesting to do this analysis for all of poling data we currently have.



Any correlation to whether the primaries/caucuses were open to Independents/Republicans or "closed"- Democrats only?


tom brady:

I understand why people do not want the Florida (or Michigan) results to count in the delegate race. but for the purposes of analyzing the demographics of candidate support, Florida is a useful data point; since neither major Democratic candidate campaigned there, and both were on the ballot, the results are valid in terms of analyzing constituency support.


Daniel T:

Tom Brady:

No, that is a bad assumption. It assumes that the fact the delegates did not count had no impact on the demographic make-up of the people who went to the polls. That is a major confounding factor that you can't dismiss with a wave of the hand. The exit poll results are not valid because the vote was not conducted under the same conditions as other polls. You can't throw an apple in with the with oranges and then claim they are all the same because they are roundish fruits.



BTW: Obama campaigned hard in Florida via proxy. Spent close to 1 and a half million $.

Only real difference, the candidates only went to raise money there, and all of them, including Obama, hinted they wanted to seat the Florida delegates (at least until they lost or won).

Statistical analysis does involve looking at the data not the preconceptions, does it not?


Annette Keller:

I think that to answer your question you have to look at religion and regionalism. There are more Jewish people in the academic and intellectual scenes of NY, NJ and FL than there are in VA, MO and IL. Because of Obama's close ties to anti "Zionist" (antisemitic?) activists in his church, and his pan to abruptly pull troops out of Iraq, quickly destabilizing the region, he probably has much less of a fan club among Jewish intellectuals than among Goys.



Obama did not campaign hard in FL. The ads shown there were part of a national ad campaign inwhich FL could not be separted out.
Also, as far as FL stats goes, Clinton won the state on early voting ballots that were cast before anyone knew who Obama was. If you read the exit polls you will see Obama did better than Clinton on the day the votes was taken.


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.