Or for the fnords.
And you think that the Democratic Party is not mostly pro corporate?
I thought someone in the pit was being sarcastic by talking about this place becoming Stormfront Lite…
Significantly less so especially once we bring the TRUMP supporters in and expel the shills of the Wall Street-Silicon Valley axis.
An angry populist party that if it ever achieved power would raise minimum wage to $75/hr and drive every job to China? Sounds delightful.
Why must minimum wage debates always be exaggerated to ridiculous levels? No, raising it to $10 or $12 or $15 is not the same as (and would not lead to) raising it to $75 or $100,000 per hour.
Ooh, fun! Can I play, too?
Here we go:
The Democrat party will make us all be gay! Religion will be outlawed! Mandatory abortions for everyone: women right away, and men after the mandatory sex change!
Let’s see, what else…
All white people lose the vote! Cars are outlawed! Veganism mandated by law!
Stop me when you think I’m getting hyperbolic.
Why not advocate an amount that can be clearly linked to the deleterious effects? Detroit is a prime example of liberalism run amok. And even that wasteland is not enough evidence. We need to go full throttle and completely price America out of global competition. Of course we’d turn a blind eye to 20-30 million economic migrants who would do the work for cash…
Detroit? With such a colossal minimum wage of $8.50 per hour (that was only increased last month)? Liberalism run amok! How will businesses ever survive if they have to pay such a massive wage?
Detroit is nowhere even close to as liberal as Norway; let’s use Norway as a prime example of liberalism run amok.
Eh, the results of that study are overblown. The biggest problem is that their results weren’t internally consistent. Their conclusion is that white names get more call backs than black names. Yet in their results Ebony received slightly more responses than Sarah. As did Jermaine over Matthew. They also don’t have an explanation for discrepancies within race. Why did Brad have a response rate of 15.9% while Neil’s was 6.6%? This effect is far more significant than the difference in the two races. Even more so for black females where Ebony was nearly 5 times as likely to get a response than Aisha.
As I understand it, they used 36 different names and sent out 5,000 resumes. Female names were disproportionately used–exclusively so for administrative/clerical jobs. Let’s say that the 18 female names were used for 3,600 resumes, or each name for 200 resumes.
If that’s the case, the difference in callback rates between Ebony and Emily (the lowest-performing white female name) comprise about 4 more callbacks for Ebony than for Emily. The difference for Jermaine over Matthew is about 1 more callback for Jermaine than Matthew might be about the same, given the lower number of male resumes–do you need me to do the math?
The point is that when you start drilling down into comparing individual white names against black names, data noise becomes very strong: four callbacks difference can easily be coincidence (Emily resumes disproportionately went to people with an internal candidate, Ebony resumes disproportionately went to employers with a sister named Ebony, etc.). The data does not support this sort of analysis; there isn’t robust enough data to draw conclusions at such a micro level.
But when you pull out–when you look at all white-sounding names vs. all black-sounding names–suddenly the effect becomes much larger. The effect becomes large enough to have some statistical power.
I don’t think the problems you’re seeing are as significant as you think.
You only addressed half of the problems I raised.
The study doesn’t report the statistical analysis, so we don’t know. But almost certainly the 10.5% call back for Ebony against a 2.2% for Aisha is statistically significant. In that case, we are looking at ~21 responses vs ~5 out of ~200. Part of writing a good paper is acknowledging and attempting to explain data that contradicts your hypothesis. In this paper the writers essentially chose to ignore it.
I’m curious if a statistician can address your claim. It’s entirely possible that the difference you describe is not statistically significant, but I’d like to see an informed post on the subject of statistical significance at this level.
In any case, the statistical power rises greatly as more names are added into the mix; variance at lower levels in no way decreases the higher-level significance.
It would negate the hypothesis of the entire paper. Say the data showed that one or more black name was not discriminated against, with statistical confidence. That would be enough to reject their hypothesis. It could not be employers discriminate against black names, because here is a black name that was not discriminated against. The authors would have to come up with a theory explaining why some black names were discriminated against but not others. That’s a huge problem because there’s no obvious explanation. The author’s “solved” this problem by essentially ignoring it.
Beyond the issues with the study, it is now 15 years old and the applications were submitted in a way that no one uses any more. It is simply very weak evidence of discrimination today.
What exactly do you see in Detroit as a prime example of liberalism?
The city started going downhill in the 1940s; in the post-war boom, there was tremendous interest in moving to the suburbs, but those were all outside of the Detroit city limits. The cities to the north grew explosively; Oakland County remains one of the ten wealthiest counties in the nation. Those left behind in Detroit were the ones without enough money and the ones unable to overcome segregation patterns (not mutually exclusive groups, BTW). Racial covenants, redlining, restrictive lending policies, and city officials actively trying to maintain “white” communities north of 8 Mile Road led to incredibly distorted housing patterns.
So you end up with an inner city populated mostly by poor minorities, surrounded by rich white suburbs. The city itself doesn’t have the tax base to maintain all of its facilities and amenities, so more people try to move up and out, further reducing the tax base, which leads to further cuts and more people wanting out. Lather, rinse, repeat.
Now factor in the city’s position in the Rust Belt, and further economic hardship wrought by the decline of the city’s manufacturing base (as well as increased automation/mechanization in the auto industry, so fewer good-paying jobs for auto workers).
Which of these are examples of liberalism?
I disagree. One would expect, given random chance, for acceptance of resumes by name to occur along a standard bell curve, right? As in, some names would have an unusually high acceptance, and some would have an unusually low acceptance. There’s not necessarily a reason for any particular name to fall at any particular point along the curve; that’s what random chance is about. Out of 36 names, one name is going to be at the extreme end of the bell curve in terms of acceptance, and one name at the other.
What’s interesting is not necessarily the percentage of acceptance of any given name; what’s interesting is the different bell curves presented by white-sounding and black-sounding names. When the curves themselves are so different, something’s up.
I think “Ebony” is more racially ambiguous than the other black female names. I’ve encountered a couple of white Ebony’s, but no white Latoyas or Tamikas. Although I have run into a white “Keisha”.
My initial guess was that there may be class bias. “Ebony” sounds more like a middle class name to my ears than a name like “Tamika”. But the average maternal educational attainment associated with “Ebony” is only a bit higher than than the one for “Tamika”. “Aisha” also sounds middle class to me (as do most names that are straight from the motherland). And yet it had the lowest call-back rate. That would strongly suggest that it’s the blackness of the name that is driving the disparities. Not class.
At any rate, I strongly disagree that acceptance of one black name refutes the hypothesis that people are biased against black names in general. People may subconsciously believe that “Jermaine” sounds like a good black guy–the kind of guy who is close to his family (especially his brothers) and knows how to sing and dance. He’ll keep us entertained, at least. But Jamal sounds like he’d be angry and militant, so let’s not call him back. Ebony sounds like she’d know how to speak proper English. Probably goes to church all the time and tells everyone to have a “blessed day”. We might be able to work with that. But Tanisha? Um, no. If we must hire the blacks, let it not be one of those. We have appearances to maintain, after all.
(I’m changing my name to Brad.)
I wouldn’t expect such a broad curve and statistically the numbers look odd. For example, Brad’s Z-score is 2.21. We’d only expect 2.7% of results to have that score or higher in a normal distribution. The other race/gender mixes are similar, though not quite as improbable.
And indeed, the numbers do show something is up. Almost certainly one of the somethings of it is measuring an actual effect. Just as certainly another one of the somethings is an error in the design or implementation of the study that lead to some names seeing unusual results.
Aisha isn’t straight from the motherland. It’s an Islamic name and to me, is most associated with India and Pakistan.
Did I say Aisha was straight from the motherland?
It is a Swahili name as well as Arabic name. A lot of Arabic names are black names, but Swahili names are also very popular in the black community.