Following the discussion here, I thought some scatterplots could be useful. I've put them below the fold so as not to bother those of you who are (justifiably, perhaps) tired of rehashing the 2008 election.
Here's the information in the second map of my earlier post, now expressed as scatterplots:
Each dot on the graphs is a county; the lines are lowess regressions. As you can see, my calculations were not perfect; in particular the formula says that Obama got less than 0% of the nonblack vote in some counties, which can't be right. Rather than correcting those points and putting them at zero, I've left them where they are as an indication of imperfection in the model.
Outside the South, Obama did best among nonblacks in counties with more blacks. Inside the South, he did best among nonblacks in counties with fewer blacks. This has got to be an urban-rural thing as much as anything else.
P.S. I like these scatterplots, but I think the maps are useful too, in particular for shooting down the story that whites in Appalachia were particularly anti-Obama.
3.23.2009
More on Obama's Vote Among Nonblacks
@ 2:49 PM
Subscribe to:
Post Comments (Atom)
54 comments
Are you kidding? Tired of rehashing the 2008 election? This is a landmark that is going to remain one of the most discussed and analyzed elections ever.
Was/is it a turning point in racial politics -- does the end of the Bradley Effect mean nothing?
Is this a critical election or a realigning election -- in terms of partisan realignment?
Does it denote a generational shift in values -- one that is likely to be accompanied by more profound changes than just who is selected to the White House?
In short, keep the data analysis coming, Andrew!
"Outside the South, Obama did best among nonblacks in counties with more blacks. Inside the South, he did best among nonblacks in counties with fewer blacks. This has got to be an urban-rural thing as much as anything else."
What about the possibility that, in the South, Republican voting among whites increases as the non-white population increases?
I'm pretty sure I read once that racism among southern whites increases as the number of non-whites in the community increases.
A must read article especially for some of the bloggers on the left who want Tim Geithner to step down
The Field: The Smart Move: Geithner Isn’t Going Anywhere
http://narcosphere.narconews.com/thefield/smart-move-geithner-isn%E2%80%99t-going-anywhere
"As you can see, my calculations were not perfect; in particular the formula says that Obama got less than 0% of the nonblack vote in some counties, which can't be right."
Anything can be right when you got bozo's like THESE.
I think that the demographic variable that indicates the "southernness" of a white voter is somehow a mix of religion (evangelical) and ancestry ("American" ancestry voters, who are mainly found in Appalachia and the other areas red in your map). In most regressions on Obama state poll averages those variables were highly significant with a negative coefficient. In Appalachia, the Black Vote was not big enough to cancel out the effect of those voters, so it shifted Republican, while in states like Georgia the increased African-American Turnout simply drowned the effect. But your map shows that it was there.
Andrew--- thanks for bringing back the old 538 feeling. In the moment, I prefer your commentary to that of Nate and Sean, by far.
Apparently the tendency for racism among whites to increase as the non-white population increases (which I alluded to in a previous comment) is known as the "racial threat hypothesis".
@Andrew: thanks for testing the regional differences in relation between %black and Obama vote.
I would surmise that you can specify all of this with an interaction effect between South and %black. But to deal with the apparent "nonlogical" results I'd take a good look at two things.
First, there are different ways to define the South -- old Confederacy, south of Mason-Dixon line. Also what about increasingly gentrified areas such as VA and a lot of NC?
Second, I'd explore urban and rural differences (e.g., %urban in county) directly in the model. In that connection, for the most part Black=Urban in the "non-South," whereas there's a much more substantial rural Black population in the South. So this takes some attention to how you specify the model.
In any case, keep it coming!
Would in be naughty of me to criticize the graphs? I hate that the top two have so much white space: it's pretty much impossible to see any detail in the West, and difficult in the others (the positive skew doesn't help. Would a logit transformation make some sense?). Would it look better with 1 column of 4 plots?
Also, why is the Northeast in the Southwest of the plot?
Overall, it might have been better to do the whole thing as a table.
*ducks*
I'm betting a lot of the positive correlation you're seeing in non-south area is explained by the urban rural split. Cities tend to both be more liberal as a whole, and have a higher percentage of blacks than rural areas outside the south.
"southerness"
I like this word,so I was looking for an example and I find out this:
"Jindal's Mockery Of Volcano Monitoring Money Only Looking Dumber After Redoubt Blows"
http://tpmmuckraker.talkingpointsmemo.com/2009/03/jindals_mockery_of_volcano_monitoring_money_only_lo.php?ref=fp1
man,ignorance is a bliss...
bye.
oh,I almost forgot:
Very interesting analysis Sir Andrew.
:)
@Adam Sharp
Do you realize that there are many reasons (I could not possibly hope to list them over the course of my life) why someone would not want to vote for Obama that have nothing to do with race?
Many Christians and Evangelicals were repulsed by his pro-abortion stance. That alone was enough for them to vote against him, race wasn't an issue.
And I just cannot get over people talking about "white racism" in the 2008 election when over 95% of blacks voted for the black candidate.
John:
Maybe we're still talking about white racism because blacks will still vote in 90+ numbers for white candidates, even losing white candidates. But white voters, as a group, seem to have a tougher time voting for a black candidate with wide popularity.
As interesting as this is, I'd be even more interested if you could do an overlay with income--from the demographics as I understand them, as the percentage of non-whites in a community rises, the average household income in the community falls--so which wins out in determining voting in the south? Income or "racial threat"?
Additionally, (not to criticize) p values or R values would give some context to the models for those of us who like numbers but aren't much good with graphs.
Wide popularity right guy won 55% of the popular vote
He won 95% of Blacks
He won 2/3 of Hispanics
He narrowly lost to Whites
Thats not wide popularity thats wide minority popularity
P.S. This is nothing new its been this way for years. And the Rs are in trouble as the country continues to diverse
And like I have been saying it wasn't just racial
Obama won 2/3 of young voters
and he tied among high income voters
I'd like to see these same studies done for over 45 voters. I wonder if Appalachia would light up bright red then.
Nova Middle Man:
You're proving my point: whites seem to be the odd group out in that a majority of them didn't vote for Obama. That's not "wide minority popularity" (see also the preference of young voters you cite), that's "a specific issue with whites." What might that issue be?
Also, again, blacks have voted in overwhelming numbers for white candidates (see Kerry). And before you say, "well, there were no black candidates in that election," check out the percentages Michael Steele got.
@ jslater
You do realize that whites typically vote for Republicans, right? Just like blacks usually vote for Democrats, right?
In the Democratic Primary, in which both candidates were liberal Democrats, over 90% blacks still voted for Obama. What do you make of that?
If you want to determine who votes along racial lines, you MUST control for party.
Let me say first of all. I think we agree more than we disagree.
I think the whole racial thing on both sides is relatively small. I also appreciate the maturity up to this point. Its tought to talk about this subject.
The tough thing about this election was it was so close all it took was a little bit of movement to impact the result. It also makes it very difficult to do analysis because of how close everything was and you obviously don't state your race/income/religion when you actually vote
With that being said there was a racial component to this race. Its not just the percentages although +8% isn't anything to laugh at. More importantly its the actual turnout. I think we can all agree African Americans turned out in much larger numbers due to the race of the candidate.
Obama won North Carolina and Virginia because of his race period.
From Wikipedia
"African American turnout increased from 11.1% of the electorate in 2004 to 13.0% in 2008.[124] According to exit polls, over 95% of African Americans voted for Barack Obama. This played a critical role in southern states such as North Carolina. 95% of North Carolina's registered African American voters turned out, as opposed to 69% of North Carolinians in general, with Obama carrying an unprecedented 100% of African American females and African Americans age 18 to 29, according to exit polling.[125] This was the case in Virginia as well where much higher turnout among African Americans propelled Obama to victory in the former Republican stronghold.[citation needed] Even in southern states where Obama was unsuccessful, such as Georgia, South Carolina, and Mississippi, due to large African American turnout he was much more competitive than John Kerry in 2004.[citation needed]"
At this point there is no more data. I can make a racial anecdotal argument for Ohio.
Obama and his supporters worked their tails off to win in Indiana and Colorado
The hispanic issue helped in Nevada and Florida and to a certain extent Colorado as well.
Bottom line it was close so close that race played an important part in the outcome.
I forgot the other part of my argument. Many people don't realize how hard Bush worked for the hispanic vote. He was fairly popular among hispanics.
McCain failed to capitilize on this even with his moderate immigration policies. The hispanic vote reverted back to the traditional D/R party split and all the gains Bush made were lost.
Great post.
Can we see the counties where Obama recieved less than 5% of the non-black vote? It would be an interesting map. That might be the true map to the racists!
John:
Sure, we should adjust for party. Republicans normally vote for whites. But that's in part -- note I said "part" -- a legacy of, among other things, the southern strategy and the fact that the Republicans run few if any black candidates. Why do you think that is?
My broader point was that blacks will vote for a liberal/popular/Dem. white candidate more, it appears, than whites will vote for a black candidate, even if, like Steele, Keyes, or Watts, the candidates are conservative Republicans.
nova_middle_man:
I appreciate the civil tone too, and I also think we may not disagree too much. Of course enthusiasm among blacks (and other minorities) was real and yes, it helped Obama. I'm just saying that he needed that because a not-insignificant subset of whites seemed unable to vote for Obama because he was black.
Now, we could debate whether or not is it exactly morally equal, given the country's history and current record regarding black federal office holders, for black folks to give an extra push for a black presidential candidate because he's black, as opposed to white folks giving an extra push to a white presidential candidate because he's white. But that might get into the touchier stuff you refer to.
Oops, obviously I should have typed "whites normally vote for Republicans," not "Republicans normally vote for whites."
@ jslater
Can you please address the point I made about black voters in the Democratic Primary? Why did 90% of them vote for the relatively unknown black candidate over the incredibly popular (among blacks) white candidate?
Just a quick comment on the coloring of the graphs: It's a bit difficult to follow the blue line in a heavily populated field of black dots. I'd recommend using a higher contrast set of colors. Replacing the blue with red, green, or magenta would make for much more readable graphs.
> Obama won North Carolina and Virginia because of his race period.
Obama won election because, when contrasted with his opponents, he was substance. He walked, talked, and projected like a President. He put in practice leadership skills that outstripped the others Dems and the best the Republicans could muster.
The clusterf*ck of Bush's Presidency was no small measure of help either, and that his primary Dem opponent was a woman was likewise extremely helpful. (( I've long said that the US, unlike say Britain, was more likely to elect a non-white than one without testicles. )) But in the end he won on merit. He very decidedly did NOT place the "race card". Because he knew damn well that a black man playing the "race card" = lose.
John: First, I don't know why any particular black voted for Obama over Hillary in the primaries. I know why I did, but I'm not black. I understand you are suggesting this is near-jerk black racism, but that doesn't explain the point I've made that you keep dodging: blacks vote for whites more than whites vote for blacks, even after adjusting for party and political persuasion.
Jslater, that is absolutely false.
Most minority-majority districts have minority elected officials. Furthermore, most districts in which minorities constitute nearly a majority of the population still have minority representatives. In fact, there was a recent debate over this issue and its relationship to the Voting Rights Act, an act which sought to create minority-majority districts because the courts knew that minorities voted monolithically for the minority candidate.
Why is so hard for liberals to admit that blacks are more racists than whites? They are far more homophobic; why can't they be more racist?
I get the impression that white liberals feel that any criticism of blacks, women, or Muslims is inherently bigoted. That's simply absurd. Every group deserves some criticism, and I know you will hate to hear this, but blacks deserve criticism more than whites.
Edit: I meant to say "Why is it so hard..."
John:
No, what I said isn't false, and apparently you haven't been paying attention. Blacks vote for liberal white dems more than whites vote for conservative, black Republicans (see campaigns of Steele, Keyes, and Watts).
You can criticize whatever group you want to. But get your facts right.
> Why is so hard for liberals to admit that blacks are more racists than whites?
Why is it so hard for you to understand why that sentiment could exist? Why do you think Obama was able to get past it? Because he didn't grow up in an environment with the venom pounded into him. Venom that then later reflects out in an ugly manner, like it does with Jesse Jackson and Al Sharpton.
It isn't pleasing to look into the mirror, is it?
Eyeballing these scatterplots, it look to me as if, except for the south, the best-fit line is being driven highly by a very, very few points, and that if those were removed or had significantly different results, the lines might have been pointing in almost the opposite direction. I'm betting the confidence intervals on those lines is HUGE. Am I right?
This is why I cannot stand arguing with liberals about anything race-related; they argue based on ideology, not on facts.
For instance, how many liberals here believe that blacks and whites are equally intelligent? I would wager everyone.
Now all of the empirical evidence (IQ tests, SATs, brain MRIs, genetics, etc) indicates that blacks and whites are not equally intelligent. Are any of you willing to reconsider your argument? Of course not. That would be "racist."
I know race and intelligence is a little off-topic, but I merely want to display how the arrant fear of being labeled "racist" prevents any meaningful discussion about race with liberals. There are only three possible outcomes that a liberal wants to hear: Blacks and whites are identical in everything but skin color; the racial achievement gap is due to discrimination against blacks; and blacks are never racist.
> Now all of the empirical evidence (IQ tests, SATs, brain MRIs, genetics, etc) indicates that blacks and whites are not equally intelligent.
You mean the ton of scientifically discredited evidence? Most use of IQ these days is recognized as outright quackery, malicious or not.
But of course there was some selective processes going on due to the slave trade, and then the brutal discrimination coupled with ingrained economic disadvantages, that followed that would have some influence the black population within the US.
Similar to how Christian european society's views on money lending and discrimination against Jews influenced that population over the course of centuries (in a way that we still see now).
But that has little to do with black = less intelligent.
Dwight,
IQ tests are actually extremely valuable. The Army has been using them for nearly a century. Furthermore, they very accurately predict success in school and work. Saying that IQ tests are "outright quackery" shows severe naivety on your part.
If I interpret your post correctly, you are saying that slavers selected less intelligent blacks to ship over to America. Why then are black Americans significantly more intelligent (over one standard deviation higher) than Africans?
The reason is that African Americans are about 20-25% white. The white genes made them smarter, and scientists are absolutely sure of this. The only reason this is new to you is because it is so politically incorrect and because the media have hid it from the public.
Dwight,
Check your facts man. African American turnout was through the roof in North Carolina and Virginia. Without that turnout Obama loses.
I have been saving this article for a while. This is my closing argument as it were.
I really wish there were more stats availalbe. But it is looking more and more likely that Ohio and Florida were also greatly impacted by the higher black turnout.
See for yourself. Black turnout impacted many congressional races as well.
http://dyn.politico.com/printstory.cfm?uuid=8E09EF5B-18FE-70B2-A87392C1847C3EEC
These races are prime targets in 2010 when blacks will have less of a reason to show up without Obama on the ticket
“African-American turnout at record levels was a necessary precondition for at least a dozen Democratic House candidates,” said Cook Political Report House analyst David Wasserman.
Bright’s district is the best example of the Obama effect. While Alabama’s 2nd District is one of the most Republican seats in the country, nearly 30 percent of the population is black — and the overwhelming majority of them voted for Obama and down-ballot Democratic candidates.
According to Bright’s pollster, John Anzalone, black turnout in the district usually accounts for about 23 percent of the electorate during presidential elections, but this time it was at least 26 percent — if not much higher.
“No one’s going to doubt black turnout was a big factor in Bright’s victory,” Anzalone said.
Anzalone also credited Bright’s popularity among culturally conservative whites, a group with whom he performed better than past Democratic nominees. Bright carried his majority-white home base in Montgomery County by nearly a 2-1 ratio over Love, 65 percent to 35 percent.
Democrats also benefited in the two Southern states Obama carried: Virginia and North Carolina. In Virginia, two first-time Democratic candidates appear to have unseated Republican incumbents on the heels of heavy black turnout.
Democrat Tom Perriello is declaring victory after holding a 745-vote lead over GOP Rep. Virgil H. Goode in a Southside Virginia district that was home to segregationist Sen. Harry Byrd’s political base for decades. The spike in turnout in heavily African-American cities like Martinsville and Danville played a key role in Perriello’s apparent margin of victory.
Perriello comfortably carried those two cities this year — after Goode won them by double digits in 2004 and 2006. In Danville, where Obama made an August campaign stop, Goode won 58 percent of the vote in 2004; that number dropped precipitously to 42 percent this year.
It was the same story in Virginia’s Tidewater-based 2nd District, where Democrat Glenn Nye defeated Republican Rep. Thelma Drake by a 4-point margin. African-Americans account for about 21 percent of the population in the district, and black turnout was up across the district.
In North Carolina’s 8th District, Democrat Larry Kissell turned a 329-vote loss to GOP Rep. Robin Hayes in 2006 into a resounding 30,000-vote victory two years later. The district, which spans the Carolina Piedmont, includes urban pockets around the Charlotte area with large concentrations of African-American voters.
The state’s Democratic Senate nominee, Kay Hagan, also cruised to a 9-point victory over Republican Sen. Elizabeth Dole.
“These were candidates running in states Obama targeted at the presidential level. They reaped some of the biggest benefits because they shared volunteers, they shared offices,” Wasserman said. “There was a coordinated Democratic effort on the ground in voter registration and persuasion.”
Heavy black turnout also affected the outcomes in two contested Georgia races. In the 8th District, which is 32 percent black, Democratic Rep. Jim Marshall won a comfortable 14-point victory against a highly touted Republican challenger — despite facing flak over his support for the Wall Street bailout. In 2006, Marshall won reelection by fewer than 2,000 votes.
In the Senate contest, Republican Sen. Saxby Chambliss was held to less than 50 percent of the vote, forcing him into a Dec. 2 runoff against Democrat Jim Martin.
“The Obama factor was a very strong reason I’m heading into a runoff,” Chambliss said, noting that black voters normally make up 27 percent of the Georgia electorate, while this year they were about 31 percent.
The spike in African-American turnout for Obama wasn’t limited to the South. It also doomed veteran Republican Rep. Christopher Shays in an affluent, predominantly white district in Manhattan’s Connecticut suburbs. In the past, Shays had performed credibly for a Republican in the urban part of the district — the city of Bridgeport — where he won by more than 30 percent against his 2004 and 2006 rival, Diane Farrell.
This year, Shays could muster only 18 percent of the vote in Bridgeport against Democrat Jim Himes, which was the difference between winning an 11th term and now heading into an early retirement.
In Ohio’s Cincinnati-based 1st District, GOP Rep. Steve Chabot had been leading Democrat Steve Driehaus throughout much of Election Night until the heavily African-American precincts in the city began reporting late in the evening. Those numbers turned a double-digit Chabot lead into a 3-point Driehaus victory.
The surge in African-American votes also may have helped the Democratic victor, Frank Kratovil, in a close contest on Maryland’s Eastern Shore. While only 11 percent of Maryland’s 1st District is African-American, Obama and Kratovil racked up huge margins of victory in precincts with large black populations.
Kratovil won the race over Republican state Sen. Andy Harris by about 2,000 votes — less than 1 percentage point.
“Make no mistake, Kratovil owes his election to Obama and the African-American community,” Edward Lee, Worcester County National Association for the Advancement of Colored People president, told the Associated Press.
> IQ tests are actually extremely valuable.
As a meaningful measure of general intelligence? Quackery. Flat out quackery.
> you are saying that slavers selected less intelligent blacks to ship over to America
I'm suggesting that the process would tend to bias towards certain physical capabilities and tolerances. For various reasons (just surviving the trip over for example, the conditions were hideous).
> The white genes made them smarter, and scientists are absolutely sure of this. The only reason this is new to you is because it is so politically incorrect and because the media have hid it from the public.
Hid from the public....with the aid of alien technology gained from UFOs? O_o OK, I'll bite. Links to peer-reviewed materials?
Dwight, I'm glad you are willing to consider other arguments.
http://www.news-medical.net/?id=9530
http://cat.inist.fr/?aModele=afficheN&cpsidt=14091763
http://www.huppi.com/kangaroo/L-IQgapgenetic.htm
http://www.innovations-report.de/html/berichte/studien/bericht-43536.html
http://www.jbhe.com/features/53_SAT.html
http://www.nytimes.com/2005/09/08/science/08cnd-brain.html?_r=1&ex=1283832
These are only a few studies available on the internet. There are many, many more if you want to read more. There are also some fantastic books such as "The Bell Curve" and "IQ and the Wealth of Nations."
> Check your facts man. African American turnout was through the roof in North Carolina and Virginia. Without that turnout Obama loses.
Go back and read what I wrote. Now what would the EC vote with NC and VA flipped? Yeah.
You also have to keep in mind that Obama's organization went out and tapped a lot of new voters. Because of historically black turn-out was quite low that represented a huge source for them. Along with the youth vote.
When your party was already pulling 88% of a demographic guess how that's going to come out?
@John, I pulled a semi-random item out of that list:
http://www.jbhe.com/features/53_SAT.html
I assumed it talked about SAT scores and I was curious about the SAT because I've not really had much exposure to the test itself (have never written it, haven't read a lot on it). Tell me, have you read it fully and what is your take-away from it?
Because I don't see it jibing with your posts.
Andrew,
Those are great scatterplots! And thanks for not fitting straight lines to them like Nate would. :o)
John,
Where is this magical land where things like poor nutrition and poverty and other socioeconomic evils have been eliminated and replaced with an egalitarian playing field? I'm dying to know.
woo hoo Dwight is back. Sorry no more troll food for you today bud. I can always tell I am right when I get a comment from you. This is fun :-)
@Dwight
I mainly posted the article to show that whites score better than blacks at all income levels. I did not agree with the analysis.
Whites from families that make less than $10,000/year score higher on the SAT than blacks from families that make more than $100,000/year. The SAT is essentially an IQ test, and in fact, the SAT correlates as well with an IQ test as it does with a retaking of the SAT.
@Space
Obviously when researchers look at the data for race and intelligence, they control for variables like poverty, malnutrition, etc. They compare blacks and whites of equal socio-economic statuses and have found that poverty or education does NOT cause the discrepancy. Genetic differences are the only explanation, which makes sense of course since race is determined by genes. I have yet to have someone explain to me why all races should be expected to have equal intelligence levels. According to Nobel Prize winner James Watson, two groups of humans that have been geographically isolated for thousands of years should not be expected to have equal cognitive abilities. The data thoroughly support his argument.
> I mainly posted the article to show that whites score better than blacks at all income levels. I did not agree with the analysis.
Well that's convenient for your argument, isn't it. Cherry picking factiods isn't exactly a convincing argument. *rolls eyes* Oh but it's all a vast media conspiracy keeping this hush, hush! Don't look now, your tin foil hat is shining through.
> woo hoo Dwight is back.
I went somewhere?
> Sorry no more troll food for you today bud.
While I must admit to enjoying chewing up and spitting/shitting out fools and trolls alike, yes I suppose you might call trolls "food" to me in a peculiar way, I would certainly appreciate if you'd STFU. Thanks! ;)
> I can always tell I am right when I get a comment from you.
I suppose it would make sense that you'd the impression you are "right" when you tragically fail at basic reading comprehension. :(
@Dwight
Conspiracy theory? Give me a break.
It is true that most major news networks would never cover such a story. Can you imagine the New York Times printing the headline, "Blacks Genetically less Intelligent than Whites." Just imagine the backlash from every crazed radical egalitarian.
And don't even get me started on universities. Who could ever forget when Harvard fired its president, Larry Summers, for saying that men *might* be better than women at math. College professors are the ones who perpetuate this nonsense, and by doing so, they hurt black people.
The best thing for race relations in this country is to accept that whites and blacks are not equally intelligent and plan our policies accordingly.
Andrew, these scatterplots are cool but not quite as useful as they could be. With almost all points clustering in the 1-10% black range, the plots are hard to read. Consider using a logarithmic scale for percentage of blacks per county.
> Conspiracy theory? Give me a break.
You then go on to talk about how some many won't talk about it a concerted effort to hush anyone that tries to. Do you know what a conspiracy is? You are clearly calling out what you see as a conspiracy of silence.
> The best thing for race relations in this country is to accept that whites and blacks are not equally intelligent and plan our policies accordingly.
Based on what? Your cherry picking of factiods? I even let you pick the URLs, and the first one I visit is a total miss for you. Ironically I suspect far more-so than you seem to realize. That article was dropping parts of the SAT scores too, the parts showing that the gap is closing.
> ... and plan our policies accordingly.
I am curious exactly what you envision this planning would involve. What policies and in which way.
am i the only one seeing something phallic about these plots?
@John
Obama isn't "pro abortion." He is "pro choice." Big difference . . .
@John
> The best thing for race relations in this country is to accept that whites and blacks are not equally intelligent and plan our policies accordingly.
Uh we already did that...last I remember it didn't turn out so well.
Man thank God those dunderhead Conservatives in the Confederacy lost the Civil War and every other large movement in our countries history.
If not John's ideas here might still be relevant to policy. Now that is ignorance I can't believe in...
Post a Comment