Open Thread and Link Farm, Frozen Like My Cold Dead Heart Edition

  1. Atheism and the Basis of Morality (pdf).The author argues that believing in an all-powerful, all-wide omnipresent, all-knowing God is incompatible with common morality, such as the idea that we shouldn’t permit a child to suffer terribly if we have a choice.

  2. Giving housing to the homeless is three times cheaper than leaving them on the streets – Vox

    Multiple studies have pointed to this same conclusion: It ends up costing cities less to provide housing for the homeless, then not. But this idea doesn’t even seem to be on the map at all, except among the left side of the left.
  3. Asking the Wrong Questions: It’s Easy to Be a Saint in Paradise: Thoughts on The Good Place’s Third Season
    wonderful and thoughtful blog post on the ethics of The Good Place, by Abigail Nussbaum. Thanks to Mandolin for the link.
  4. Scott Satens provides highlights of the results of the recent Basic Income experiment in Iceland.
    Interestingly, people on Basic Income got slightly less assistance than the control group on welfare – but reported higher levels of happiness.
  5. Americans are becoming more socially isolated, but they’re not feeling lonelier
  6. Self-Styled Free Speech Advocate Dave Rubin Praises Jair Bolsonaro For Ridding Schools Of ‘The SJW Stuff’ – Angry White Men
  7. Donald Trump handed a chance to supercharge voter suppression in 2020.
    I don’t think the decision was necessarily legally wrong. But it still frees up the RNC to step up voter suppression efforts – specifically, sending “security” personnel to go stand in front of minority-heavy voting locations.
  8. Elizabeth Warren Apologizes to Cherokee Nation in Private For DNA Test
    Some Cherokee Nation members are saying that the apology should have been public.
  9. The Misguided Focus on 1619 as the Beginning of Slavery in the U.S. Damages Our Understanding of American History | History | Smithsonian
    The story is more complex, and goes back further, than some people think.
  10. The Destruction of Black Wall Street – by Chelsea Saunders
    A short comic about the 1921 white riot that destroyed a prosperous Black community.”
  11. 16 Black Moms Are Getting a Basic Income in Mississippi. Here’s How It’s Working. – Rewire.News
    It’s a pilot program; if it goes well, the plan is to expand it.
  12. How Big Is the Male-Female Wage Gap, Really? – The Atlantic
    There is no “real” answer to this question, because – even when well-done – the answer to “how big is the wage gap” is always dependent on which wage gap is being measured. This is about measuring the wage gap over the course of fifteen years; measured this way, the answer is 50%.
  13. Pelosi Aide Tells Insurance Executives Not to Worry About “Medicare for All”
    Aside from the “Pelosi is anti M4A” part of things (which is not a surprise), it’s interesting to see what they’re thinking about Obamacare and pharmaceutical prices.
  14. 3 philosophers set up a booth on a street corner – here’s what people asked
  15. Are You a Woman Traveling Alone? Marriott Might Be Watching You. – Reason.com
    As part of an anti-trafficking initiative, hotels, airlines, etc, are being told to be suspicious of things like single women and interracial families.
  16. Cops Say Cindy McCain Didn’t Catch Toddler Trafficker at Airport: Reason
    “I went over the police and told them what I saw and they went over and questioned her and, by God, she was trafficking that kid. She was waiting for the guy who bought the child to get off an airplane.” That’s McCain’s version of the story; the truth appears to be, McCain called the police on a perfectly innocent mom who was with her different-race child.
  17. This filmmaker sat down with neo-Nazis and jihadists. Here’s what she learned. – Vox
  18. Restorative Justice | Thing of Things
    “There are some things you are entitled to that are completely non-negotiable, no matter how bad a person you are…. You have a right not to be tortured. You have a right not to be assaulted or killed, except when necessary to defend others. You have a right to food and water and shelter. You have a right to human interaction (but not to force unwilling people to interact with you, and that sometimes means sufficiently disliked people are doomed to loneliness– but it is a tragedy, every time).”
  19. U.S. Economy: Higher Minimum Wages Haven’t Increased Unemployment – Bloomberg
    The evidence is now clear: Either raising the minimum wage doesn’t increase unemployment, or it increases unemployment by such a small degree that it can’t be reliably measured.
  20. I might simply link to Henry Farrell’s response to people who want to engage him on race and IQ, next time someone wants to engage me on the subject. (A reminder: arguing for so-called “scientific racism” is not allowed on “Alas.”)
  21. Venomous yellow scorpions are moving into Brazil’s big cities – and the infestation may be unstoppable
    A lot of causes – including sanitation issues and global warming – have combined to make a possibly insolvable problem.
  22. Causation Fallacy 2.0: Revisiting the Myth and Math of Affirmative Action.
    Even if Harvard and other schools that are the focus of current lawsuits stopped admitting Black and Latinx students entirely, that would not significantly improve the odds of admission for White and Asian students. (Journal article, sci-hub link).
  23. Is Harvard Really Biased Against Asian American Students?
    Asian students are correct to think they’re being discriminated against in admissions. But that discrimination isn’t due to formal AA programs benefiting minorities, but due to informal racism for the benefit of white applicants. “These findings suggest that the same constellation of grades, activities, awards, essays, and test scores are interpreted as intellectual curiosity and academic excellence when presented by white applicants, but interpreted as evidence of an unimaginative applicant who is “booksmart and one-dimensional” when submitted by a student who is Asian American.”
  24. This Nancy strip is brilliant.
  25. Carol Anderson on Republican voter suppression – Vox
  26. Latino Turnout Surged. Then Texas Questioned 98,000 Voters’ Citizenship. | HuffPost
    We are increasingly dividing into the pro-democracy and anti-democracy parties.
  27. Can bees do math? Yes – new research shows they can be taught to add and subtract
    But can they be trained to do my taxes?

This entry posted in Link farms. Bookmark the permalink. 

31 Responses to Open Thread and Link Farm, Frozen Like My Cold Dead Heart Edition

  1. 1
    nobody.really says:

    …believing in an all-powerful, all-wide, all-knowing God is incompatible with common morality….

    Fat-shaming on Alas, a Blog?

    (Well, if you start from the premise that God is everywhere….)

  2. 3
    lurker23 says:

    Causation Fallacy 2.0: Revisiting the Myth and Math of Affirmative Action.
    Even if Harvard and other schools that are the focus of current lawsuits stopped admitting Black and Latinx students entirely, that would not significantly improve the odds of admission for White and Asian students.

    i read the article and i think they make two mistakes, one big and one less big. i will explain why.

    the big is funny because it is a very obvious mistake.

    if you want to talk about “how a change affects white and asian people” then you need to look at “white and asian people before the change” and “white and asian people after the change.”

    but what THIS paper does is to look at “*all* people before the change” and compare them to “white and asian people after the change.” the groups are not the same! and because the group that is included in “all people” has a better admission than white and asian people then the study will show an increase that is too small!

    i do not know why they would do that, it is a very simple mistake. but that makes the paper not right.

    the second part is not a mistake exactly but they make a bad choice to how they show the math. to show you what i mean i will give you a test:

    a number changes from 1% to 1.5%. ask yourself “is that alot of change?” and think what your answer was.

    now a number changes from 50% to 75%, ask yourself “is that alot of change?” and think what your answer was.

    now, think of why your answers are the same or different and why. i think that is fun!

    of course you all can see that the “rate” change is the same for both of them, they both got 50% bigger. but alot of people do not really think of 1%-1.5% as a change at all because alot of people do not think about rates when the numbers are that little.

    so when you report alot of LOW numbers and if you do NOT also talk about “rate” changes (like 1 to 2 percent is a double!) then i think alot of people may not know how big the change is.

    but you can choose! remember the paper says “not significant” and here are the numbers from the paper (page 18) to show an example.

    HARVARD
    5.84% “all admissions” (which should be “white and asian” like i said so it should be lower but it is what they show.)
    6.84% whites and asians only.
    1% INCREASE (should be higher, like i said)
    17.1% CHANGE (should be higher)

    ANN ARBOR
    2.82% INCREASE
    8.5% CHANGE

    CHAPEL HILL
    4.91% INCREASE
    17.8% CHANGE

    AUSTIN
    10.91% INCREASE
    27.1% CHANGE

    so for me i look at the changes and i think that is alot of change. if i were trying to get into harvard i would be very happy to have a 17.8% increase in my changes even if it was still very hard to do. i would not say “not significant change” at all. and of course the change is going to get bigger not smaller if you do the math right.

    anyway Ampersand i think maybe you should make that more clear in your post :)

  3. 4
    lurker23 says:

    the paper only reports the INCREASE and not the CHANGE. like for Harvard they only say “it goes up by 1%” but they do not say “going up by 1% means a 17.8% increase in your chances.”

    when i see that choice to not-report and i also see the funny stuff with “all admits” compared to white and asian admits, i start to wonder if the paper is really very honest.

  4. 5
    LimitsOfLanguage says:

    2) The article compares the cost associated with the homeless with the cost of providing housing, based on the assumption that homelessness is the cause of the problems of the homeless. This is false. Homeless people frequently have underlying problems (that often caused the homelessness in the first place), like mental issues or addictions. If you give them a house, they will still have more police contact, will be jailed more often, will need more psychiatric and healthcare, etc, on average. So the evidence that is presented doesn’t support the claim.

    4) The control group was not a proper control group, because getting additional benefits was dependent on being on traditional welfare. So the people in the experiment were different from the people in the control group. Secondly, the main claim of UBI advocates seems to be that it greatly reduces the poverty trap, where working more means losing benefits, making it not worth it to work yourself out of poverty. This experiment found no increase in employment, suggesting that the experiment failed in its aim. It reduced support for an UBI in Finland.

    12) The article falsely (and hyperbolically) claims that women struggle to get hired, which they ‘show’ with cherry picked evidence. Studies sometimes find hiring discrimination against women and sometimes in favor. On average the level of gender discrimination in hiring seems low, aside from very regressive and very progressive employers (with the former discriminating against women and the latter against men). The article also fails to note the discrimination of men that I wrote about here recently: that men get punished for taking time off more than women. In general, the article is women-centric, like usual, hoping that policy changes “give women more control over their working lives,” yet never considering how men might benefit from more control over their working lives.

    22) The exact same arguments in the paper can be used to argue that the low rate of women and minorities among CEOs is no big deal…

    23) This could be a case study in how selectively presenting evidence can result in people with certain biases to draw certain conclusions. The conclusion that Ampersand draws is only valid if black applicants are judged similarly negatively as Asian applicants when being studious, which seems unlikely. After all, the stereotype is of the nerdy Asian who has very limited interests. That is anti-Asian, not pro-white, so one would expect all non-Asians to benefit from that (in a relative sense, by not being subjected to that discrimination).

    26) Yet in Europe (and Venezuela), it is ‘populists’ who fight for more democracy and more influence by the weakly represented underclass (including through referendums). I suggest looking at this through the lens of who benefits from franchising or disenfranchising certain groups. Looking at who benefits is often far more instructive than taking people’s claims to have certain principles at face value. Many ‘principles’ are instrumentative claims, not terminal values.

  5. 6
    Harlequin says:

    Lurker: I wouldn’t say that it’s straightforward which number you should use, the fractional change or the absolute change. Take your 1 to 1.5% chance of getting in. You can phrase that as “Your chance of getting in went up by half!” But you can also phrase it as “Your chance of NOT getting in is essentially unchanged!”

    Which one of those is more representative of reality? I’m not sure. Depends on what you care about.

    Like you, I’m not sure why exactly they are comparing to the admission rate for all students when they apparently had the data to do something else. (The authors say the data is publicly available online but I wasn’t able to find it; if anyone knows, and it’s not in a terrible format, I might look at it, because I’m curious now.) I doubt the difference is large, however, since most applicants and admitted students at, say, Harvard are white or Asian to begin with.

  6. 7
    Kate says:

    The article compares the cost associated with the homeless with the cost of providing housing, based on the assumption that homelessness is the cause of the problems of the homeless.

    No, it doesn’t. It assumes that people need the stability of a home before they can address their problems, and in addition to housing, it provides “a caseworker to supervise their needs.”

    Homeless people frequently have underlying problems (that often caused the homelessness in the first place), like mental issues or addictions. If you give them a house, they will still have more police contact, will be jailed more often, will need more psychiatric and healthcare, etc, on average.

    Yes, but having a home to go to and a social worker to coordinate care dramatically decreases the rates of police contact and, most crucially, very expensive emergency room admissions, for most people.
    The Vox article did not present this very well. A pilot program serving 85 chronically homeless people in Charlotte, NC found that:

    Residents of Moore Place collectively visited the emergency room, an expensive but not uncommon way homeless people access health care, 447 fewer times in the year after getting housing, the study discovered. Similarly, they spent far less time running afoul of the law, with the number of arrests dropping 78 percent.

    Lower cost. Less human misery. Win/Win.

  7. 8
    lurker23 says:

    Harlequin says:
    February 12, 2019 at 12:47 pm
    Lurker: I wouldn’t say that it’s straightforward which number you should use, the fractional change or the absolute change. Take your 1 to 1.5% chance of getting in. You can phrase that as “Your chance of getting in went up by half!” But you can also phrase it as “Your chance of NOT getting in is essentially unchanged!”

    the people who want to get into harvard are very interested in getting in and i think alot of them spend alot of time and money on things that make their chances bigger even if they are only a tiny bit bigger. for those people 17% better chances is really alot of a change, a huge difference even if they do not get in at all, so i think any article like this would be honest and at least report it.

    harvard is like winning the lottery, i think, so you can also think of a littery. i would much rather have 117 lottery tickets than 100 lottery tickts (or 117,000 tickets and not 100,000 tickets) even though the chances of winning were very very small. the people who are not willing to risk not-winning are not playing the lottery or applying to harvard.

  8. 9
    Harlequin says:

    (Edit: It took me long enough to write, step away, and then edit this comment that Kate already addressed the first part, including the exact same evidence, lol! Great minds etc)

    LoL:

    Homeless people frequently have underlying problems (that often caused the homelessness in the first place), like mental issues or addictions. If you give them a house, they will still have more police contact, will be jailed more often, will need more psychiatric and healthcare, etc, on average.

    In fact, housing-first initiatives greatly reduce incarceration rates and emergency-room use, for example. And addressing these issues is why the proposal is to provide “supportive housing”–including a caseworker to help with the other needs. So no, it’s not based on a false assumption that homelessness is the source of the problems homeless people have; it’s based on the evidence that the multiple problems many long-term homeless people have are easier and cheaper to solve with stable housing and support.

    Secondly, the main claim of UBI advocates seems to be that it greatly reduces the poverty trap, where working more means losing ben
    efits, making it not worth it to work yourself out of poverty. This experiment found no increase in employment, suggesting that the experiment failed in its aim.

    That is not the main claim of UBI advocates (I quite enjoyed the paragraph beginning “Many reasons have all been invoked in Basic Income’s favour…” at this link, but really any Google search for “reasons for UBI” will return lots of things that have nothing to do with the poverty trap–which, in any case, has been at least greatly reduced by better welfare design). And secondly, the claim I see most often is that UBI will reduce work rates (e.g. the paragraph beginning “Finally, there are many ethical problems…”, or here [which does mention the poverty trap–but the writer is not an advocate of UBI], or here which frames less employment as a positive). So no change in work hours is better than expected (at least if you want people to keep working).

  9. 10
    Harlequin says:

    lurker: it’s also worth mentioning that that is the change if there were no black or Hispanic applicants at all–not just if those applicants got a boost from affirmative action. Think of it as an unrealistic upper limit, not the size of the effect.

    I also have my usual problem that college admissions–like hiring, and other places affirmative action is an issue–is based on proxies for what you really care about, like intelligence and ability to work hard. So is a difference between SAT scores because colleges are giving a boost to unsuitable applicants based on race? Or is it a recognition that SAT scores are less sensitive to the intrinsic qualities & preparation level of certain racial minority groups, because of various correlations with discrimination and SES? Some of column A, some of column B?

    This stuff is hard. I grew up in Iowa, and I know someone who was an alumnus and did local interviews for one of the Ivies; he did them for 20 years before a single student he interviewed was accepted. I don’t think that was deliberate discrimination based on location, so much as ignorance of the culture–tall poppy syndrome means the kids are less willing to brag, for example, and lower population density means the reputations of the high schools are less known to recruiters. As I recall, this impression of mine was borne out by statistics about the states of origin for students at that particular university, but I was in college last time I looked it up so I could be wrong…

  10. 11
    Harlequin says:

    I really enjoyed the article about the Ask a Philosopher booth! Sounds like they got some interesting questions. And the end was very cute. :)

    When I was a grad student, I once did some crowd control for a long line of people waiting to get into a VR astronomy exhibit. I walked up and down the crowd, giving people the expected wait times and asking if they had any questions about astronomy. At this point, I learned that almost everyone, given the opportunity to ask anything at all about astronomy, will ask one of three questions: what are wormholes, really; what are black holes, really; and why isn’t Pluto a planet any more. I did get one fun question about the relationship between the relativity of time and the lengths of years on different planets, though.

  11. 12
    LimitsOfLanguage says:

    Kate,

    It assumes that people need the stability of a home before they can address their problems, and in addition to housing, it provides “a caseworker to supervise their needs.”

    No, the article is titled “giving housing to the homeless is three times cheaper than leaving them on the streets.” Even if we grant them a lying/clickbait title, they still claim in the article itself that giving housing is cheaper. Ampersand replicated this claim in his commentary.

    To support this claim the article tells us that $31k is spent per homeless person on “the salaries of law-enforcement officers to arrest and transport homeless individuals — largely for nonviolent offenses such as trespassing, public intoxication or sleeping in parks — as well as the cost of jail stays, emergency-room visits and hospitalization for medical and psychiatric issues.” It contrast this with: “getting each homeless person a house and a caseworker to supervise their needs would cost about $10,000 per person.”

    31k/10k = 3 times. This is how they got to “three times cheaper.”

    However, it’s only actually three times cheaper if all those $31k in costs go away if you give the homeless homes. This is basic math. If you spend $10k and the costs go down by only $10k, it’s just as cheap to give people homes (10k/10k = 1). If you spend $10k and the costs go down by $5k, it’s 50% more expensive to give people a home and case worker (5k/10k = 1/2). So you need to know how much the costs actually go down by when giving people homes and case workers, not how much is currently spent.

    It’s actually absolutely impossible for those $31k in costs to go away completely, because the public spending for policing trespassing & public intoxication, jail stays, emergency room visits and hospitalization are not $0 for the average person who lives in a home. So even if we assume that giving these people a home and case worker would make them just as (un)likely to trespass, get drunk publicly, get jailed, visit the emergency room or be hospitalized (without insurance?), that still wouldn’t make all those $31k in costs go away. As I’ve argued, it’s pretty certain that after giving them a home they will incur higher public costs than the average person who is housed.

    The Vox article did not present this very well.

    It tells a blatant falsehood. That’s not the same as making a poor argument. You are making excuses for falsehoods that match your ideology. Are you this forgiving when falsehoods are told that go against your beliefs?

    A pilot program serving 85 chronically homeless people in Charlotte, NC found that: […].

    Lower cost. Less human misery. Win/Win.

    If I look at the original story (instead of the more biased website that you linked to), it’s argued that billed healthcare went down by $1.8 million. However, uncontracted healthcare is notoriously overbilled compared to actual cost (as noted by the very study that the story is partially based on). Presumably, very little of this is paid for by the homeless and instead, it is mostly written off and the costs of that care passed on to other ER care recipients. This results in overly high ER bills. If you reduce use of the ER by non-payers, you will actually save less than seems at first glance.

    Let me demonstrate (note that the dollar amounts in the example are merely illustrative):

    Imagine that the actual average cost of ER care is $1k per visit, but that 50% of users do not pay. Then when you have 10 visitors, but only 5 payers, those 5 people have to pay double the bill, as they have to pay for those who don’t pay. So the bill per person is $2k and the total billed cost of the care for these 10 people is $20k, although the actually paid bills only add up to $10k. Also imagine that the non-ER care that the non-payers should actually use costs $500 on average.

    Then a naive view is that if we switch a homeless person from ER care to non-ER care, we will save $2k – $500 = $1500 per person and thus 5 * $1500 = $7.5k in total.

    However, if we switch all 5 of those non-paying ER users, the unpaid bills go to zero, which means that the remaining people can be billed for their actual cost, as they no longer have to pay for the non-payers. So the ER bill per person goes down to $1k, for a total of $5k. The actual saving per person is then $1k – $500 = $500 and thus 5 x $500 = $2.5k in total. That’s merely a third of the savings we earlier calculated when using inflated billing!

    So the $1.8 million in savings that is claimed by the article is higher than the real savings, perhaps by a large amount, especially since it only looked at the reduced ER bills, not the increase in bills for regular care. Of course, the article also didn’t quantify the costs in reducing policing costs.

    I don’t see any good evidence in the article or study for what the actual costs and savings were overall for this homeless facility. They don’t seem to have done the necessarily (and very complex) financial analysis.

    Also note that this particular facility does not in fact give people housing, but demands 30% of people’s income as rent. The total number of homeless in Charlotte seems to be around 600, but the article claims that only 200 qualify for this facility. So are they cherry picking the least problematic third? Also, there was a 15% dropout rate. These people went back to the streets.

    Anyway, my argument is not that this policy is not worth doing. I’m very supportive of helping the homeless, even if it costs more money than not helping them. I just want the argument for it to be made based on truth, rather than falsehood.

    If you are going to accept very poor ‘evidence’ that supports your preferred policy, then you are fundamentally no different from those who disbelieve in climate change based on very poor ‘evidence’ that supports their preferred policy. I think that if you oppose the one, you should oppose the other.

  12. 13
    LimitsOfLanguage says:

    Harlequin,

    I do agree that freedom is a common argument in favor of an UBI, although I think that many people fail to appreciate the positive freedoms resulting from capitalist disciplining (basically, capitalism forces us to do things that make other people happy enough to pay for them, so money-based trading is actually utils-trading).

    And secondly, the claim I see most often is that UBI will reduce work rates […] So no change in work hours is better than expected (at least if you want people to keep working).

    This base income was given to people without work, so there was very little opportunity for people to reduce working hours in the first place.

    They didn’t get a job less often than the control group, which can mean that work is not less attractive; or it can mean that the reduction in the poverty trap offsets the less attractiveness of jobs. If the latter is the case, people who are not subject to the poverty trap (who were not studied here) may reduce their working hours if subject to an UBI.

  13. 14
    LimitsOfLanguage says:

    As for affirmative action, you can also frame it like the applicants who filed a lawsuit did:

    Consider the example of an Asian-American applicant who is male, is not disadvantaged, and has other characteristics that result in a 25% chance of admission. Simply changing the race of this applicant to white—and leaving all his other characteristics the same—would increase his chance of admission to 36%. Changing his race to Hispanic (and leaving all other characteristics the same) would increase his chance of admission to 77%. Changing his race to African-American (again, leaving all other characteristics the same) would increase his chance of admission to 95%

    Imagine a situation where a race-based preference policy makes a white person have 95% chance to be admitted, while if he were exactly the same, but black, would have a 25% chance. Would this then be accepted as insignificant?

  14. 15
    Jeffrey Gandee says:

    I’m with LoL on the presentation of the data at link 23, it’s wildly misleading and seems designed to fool people who aren’t good at math (or reasoning at all, really).

    Worse, I’m not sure why anyone should care about differing acceptance rates for different groups, as they seem irrelevant as to whether the admissions system is fair.

    Suppose I start a charitable nonprofit, the goal of which is encourage every black student in the USA with a chance of getting into Harvard to apply there, and my organization doubles the number of applications from black students. This would result in cutting the acceptance rate nearly in half! Should Black students be angry about this? Has my nonprofit made the system less fair? I’d say “no” on both counts.

    Similarly, I’d argue the Asian students can’t use their lower acceptance rates to argue that the system is unfair, because “raw number of applicants with similar skin color” shouldn’t be a factor when deciding whether or not to admit a student.

    LoL’s way of measuring fairness at comment#14 seems right to me, and probably for most voters in the USA.

    The questions is, should the process be fair? I’m not a fan of AA for most college admissions, but my understanding is the graduation rates for black students at the Ivy’s are actually really high, in line with other groups of students ( https://www.jbhe.com/2013/11/black-student-graduation-rates-at-high-ranking-colleges-and-universities/ ). This is because everyone who makes it in to an Ivy is a good student. On the other hand, AA at mid tier and non-ivy upper-tier universities seems to be driving down graduation rates for black students who may have been better off attending schools where their classmates aren’t testing an entire standard deviation ahead of them. Maybe there should be more AA at the Ivy’s, and less at schools where AA can be shown to harm the future outcomes of black students. Then again, the differences in graduation rates at non-ivy’s may have nothing to do with talent, and more to do with other circumstances. The debate surrounding AA is complicated and politically polarized, but I think it’s easy to make a case for AA at the Ivy’s, so long as people are willing to accept an unfair application process.

  15. 16
    lurker23 says:

    Jeffrey Gandee says:
    February 13, 2019 at 6:01 am
    I’m with LoL on the presentation of the data at link 23, it’s wildly misleading and seems designed to fool people who aren’t good at math (or reasoning at all, really).

    what about me?

    also i do not know if graduation rate is really good to get all the right data. when i read things on aa i think that you should also be looking at things like how they did and how the other students in their class did and what majors they take and that sort of thing. like some schools that are not high ranked graduate alot of black doctors and engineers and physics people and other schools that are high ranked do not do that so much.

    I think it’s easy to make a case for AA at the Ivy’s, so long as people are willing to accept an unfair application process.

    if you only do aa at ivy schools or maybe top ten schools, and everyone else has to “go to a school where they can get in without aa” then you will have a very very VERY VERY big gap right under the ivy where you are really not going to have many black people in the other schools at all.

    there are maybe 4000 new students each year in the ivy league, say you are only talking about AA at the top schools that take a total of 10000 students.

    if you want 25% of them to be black you are taking almost every single black person who scored above a 1400 (there were only 2600 of them in 2018) and putting them only in those schools, so there are no smart black people left for any other schools. but you have to go way down the ranking list before you find alot of schools with an sat average below 1400. and the lower ranked schools are even bigger so they need more people.

    i saw 2018 sat scores here https://reports.collegeboard.org/pdf/2018-total-group-sat-suite-assessments-annual-report.pdf

    if you look at “who got a 1200 or more” i think that is 57% of asians who took test, 35% of whites who took test, only 8% of blacks who took test. those are bad ratios.

    if you look at 1400 or more it is even worse, only 1% of blacks who took test got at least a 1400, but 24% of asians did and 8% of whites.

  16. 17
    lurker23 says:

    to show what i mean, i did the math and in 2018 here are the people who scored at least a 1400:

    52,313 asian people
    2,633 black people
    74,466 white people

    129,412 total people

    then i did the math to ask “what percent of the high sat people are the asian black and white people?” and

    40.5% asian
    2% black
    57.5% white

    do you see the problem? there are not enough high sat black people to fill all the slots that the schools want to fill unless they use aa.

  17. 18
    Harlequin says:

    They didn’t get a job less often than the control group

    Which is what I meant by talking about comparative work hours–compared to the control group, not to the same people pre-UBI. Sorry if that wasn’t clear!

    which can mean that work is not less attractive; or it can mean that the reduction in the poverty trap offsets the less attractiveness of jobs. If the latter is the case, people who are not subject to the poverty trap (who were not studied here) may reduce their working hours if subject to an UBI.

    It’s possible–no single study is able to get at the full truth. But this is positive evidence that one of the proposed harms of UBI (disincentivization to work) is, at least, much less dangerous that portrayed by critics.

  18. 19
    Jeffrey Gandee says:

    Sorry, Lurker, I got you guys confused. It was your point on the presentation of that data that had me nodding along. I still can’t believe the author isn’t being intentionally misleading.

    AA is like most policy, it has trade offs. I suppose you make a good case for AA across the board or no AA at all. A whole bunch of priorities need to be straightened out before anyone can have an intelligent conversation, though, starting with the value add of a potential college degree in a desired major vs the value add of a different diverse campus, taking into consideration the odds of attaining said degree with and without AA. This is already a tough calculation, and we havent even considered the value add of having a fair selection process.

  19. 20
    Harlequin says:

    Imagine a situation where a race-based preference policy makes a white person have 95% chance to be admitted, while if he were exactly the same, but black, would have a 25% chance. Would this then be accepted as insignificant?

    Likely not. But that doesn’t mean the situation in the lawsuit (if it is as described) is necessarily wrong–or, rather, that the right level of difference between kids of different races should be 0. If you come from a worse school and/or a lower SES background, you had to work harder to get the same level of qualifications, and those qualifications should be judged accordingly. And we use race because we know simple measures of SES often don’t capture the full story of this background (e.g., how wealthy black kids tend to live in neighborhoods with high poverty rates even though they’re not poor themselves). It is not perfect, and absolutely upweights some people who don’t need the help and downweights people who do–after all, refugees from Burma and the kids of Google workers who immigrated from China are both “Asian.” But it’s not like a system that didn’t take this into account would be fairer–it’s overall much less fair that race-conscious admittance procedures, IMO. I think your hypothetical is unfair because I can’t imagine a situation in current US society where black people would have the kind of advantage over white people that needs to be corrected this way, not because I think any differences based on race are bad.

    There’s a separate question, again, of whether the level to which this is done is the “right” one–the one that minimizes unfairness. And how much other things are weighed, as Jeffrey says, such as desiring a diverse student body that maximizes the education of the students who are there: if the whole student body is wealthy and white, the opportunities for learning go down, so the question is not just maximizing the utility to the individual student but also to all of their classmates. Which is kind of a thorny problem to think through. I’m arguing against the idea that a fair admissions process is race-blind, which seems implied by your question, not arguing for the particular implementation at any one university–as Amp points out, there is racism against the non-academic qualifications of Asian students and that is clearly bad.

    I am absolutely saying, for example, that the average black woman with the same qualifications I had should have gotten into better schools than I did. I worked hard, but my hard work was made more effective because of things that had nothing to do with me (like: my school teaching me how to use a library, so I could learn astrophysics on my own when I was a teenager, without expending effort on figuring out how to obtain the right kind of books; the library being close enough to my house that transportation wasn’t an issue, and including books of the right content level; not being harassed for being in the university library when I went there, because I looked like the kind of person who belonged).

    This is especially relevant to the SAT scores you’re citing, lurker. SAT scores are a bad indicator of college achievement, are fairly easy to learn as a skill (especially with pricey test-prep services that wealthier parents can afford), and are known to have strong biases based on SES. You cannot use them on their own if you want an accurate result, especially for group-level proxies for ability.

    Jeffrey, I don’t have time to look up the statistics right now, but I’m pretty sure there isn’t a dropoff in graduation rates below the Ivy League–instead, there’s a pretty direct correlation between the resources of a school and their minority graduation rates. (Probably some combination of better financial aid & more support services–paperwork alone can be a barrier to students who aren’t used to dealing with bureaucracy, so better-staffed offices can help more.)

  20. 21
    Harlequin says:

    Oh, I forgot to say:

    I still can’t believe the author isn’t being intentionally misleading.

    Unfortunately, lots of people are very bad at statistics, including lots of really smart and/or well-educated people. And too often it’s not recognized as a separate discipline you should seek independent experts in. (I’m thinking of the very tragic case of Sally Clark in particular.)

  21. 22
    LimitsOfLanguage says:

    Harlequin,

    If you come from a worse school and/or a lower SES background, you had to work harder to get the same level of qualifications, and those qualifications should be judged accordingly.

    While that is true, that doesn’t mean that the gap between various ethnic groups is solely due to those causes. It seems to me that Asians do better on average in part because of cultural differences that makes them work harder. These Asian students are punished by AA for working harder, which is not justified by the reason that you give for having race-based AA.

    Ultimately, the fact that Asians do better in academia than white people and are most disadvantaged by AA is incompatible with the SJ narrative that the different success rates of ethnic groups is entirely due to oppression of people of color by the majority of white people. Yet I see people cling to these beliefs and thus act as if they are true despite the evidence against it staring them in the face.

    If you make policy based on the idea that the entire disparity between ethnic groups is being caused by discrimination and/or is fairly easily fixable, then you can have severe negative effects.

    Ill-prepared students don’t magically become capable if you let them into colleges. Requiring colleges to offer remedial education is an immense drain on them. Furthermore, it’s questionable whether they are capable of actually bringing a sufficient number of student up to par. I’ve heard quite a few horror stories by teachers who had classes where only a fairly small fraction of students was capable of learning advanced material, yet the administration forced them to not flunk these students. Many of them then decided to teach to a lower standard.

    This in turn is very negative to the students (of all races) that are capable of more, but are denied an education on par with their abilities. This then extends to society as a whole, which then doesn’t get to benefit from having students educated to a higher standard.

    We’ve also seen cases where progressives decided that disciplining in education was racist, because it was disproportionately black students who were being disciplined. Then they removed disciplining for these students, even though it was a response to disruptive behavior, which then became unchecked, ruining the educational environment for all (and ironically, making the problem of certain schools giving people a poor start even worse).

    This is a common error in progressive policies: addressing the symptoms rather than a cause, in a way that just makes the problems worse (in the previous example, the real thing what needs to be fixed is the reason why black students become more disruptive).

    But it’s not like a system that didn’t take this into account would be fairer–it’s overall much less fair than race-conscious admittance procedures, IMO.

    That depends on whether one thinks that institutional racism is ever justified or whether one rejects it on deontological grounds. I do the latter, in part because it seems ripe for misuse and conflicts over such policies inevitably generate racial resentment. I prefer not to breed white (or black, Asian, etc) nationalists.

    Even if one accepts the idea that institutional racism is a valid tool to offset societal racism, it suffers severely from the issue that societal racism is not distributed evenly. Affirmative action helps those who are least effected by societal racism the most, which means that it doesn’t actually offset societal racism on the individual level. Now, my experience is that most SJ advocates are collectivists, who are fine with policies that are unfair to individuals, as long as the unfair benefits to people in a group offset the unfair deficits to people in a group. However, I reject such thinking.

    Aside from the unfairness to individuals, an issue is also that SJ ideologues and progressives in general have a strong tendency to overestimate the level of discrimination against those they deem to be victims & underestimate the agency of those people (and underestimate the discrimination against those they deem to be oppressors and overestimate their agency).

    So I think that your side has a chronic tendency to do way too much ‘benevolent’ discrimination, especially since root causes are often unaddressed in favor of turning knobs that can’t offset societal unfairness. Then when these interventions don’t remove the societal unfairness, which they couldn’t be expected to do since they don’t address the root cause, the tendency is to turn the knob harder, with little care for the people in the outgroup who are hurt in the process.

    I think your hypothetical is unfair because I can’t imagine a situation in current US society where black people would have the kind of advantage over white people that needs to be corrected this way, not because I think any differences based on race are bad.

    Just replace black people with women and white people with men, then. Affirmative action and other specific favors for female students is still way, way more common than favors for male students, despite female students doing much better on the whole. Yet somehow the very thing that we are not allowed to do for black people or women, hold them (partially) responsible for poor outcomes, is the common and usually sole reply to the issue of male students not doing that well.

    Note that my suggestion is a moderate position, where we recognize a limited level of agency for all.

    if the whole student body is wealthy and white, the opportunities for learning go down

    If you abolish AA, the whole student body is not going to be wealthy and white. That is certainly not going to be the case if you replace AA with a SES-based policy. You are engaging in sophistry with such an absurd claim.

    Secondly, the opportunities for learning may not meaningfully be reduced in many cases. If I am studying math, then there is no black or white math, no rich or poor math. If the goal is to teach people certain civic lessons, then it is far from given that this is best taught by having people of different races interact more, rather than teach it explicitly. Note that Robert Putnam famously found that racial diversity atomizes people and destroys trust in neighbors and institutions. So AA may teach things that aren’t helpful. If you let in black people that are less capable, then student’s lived experience will be that black students are less capable than white students on average. Is this a lesson that you want to teach people?

    not arguing for the particular implementation at any one university–as Amp points out, there is racism against the non-academic qualifications of Asian students and that is clearly bad.

    Affirmative action is discrimination against people by race based on non-academic qualifications. That’s what it is!

    You can’t favor a policy and then run away from its negative consequences that are unavoidably part of that policy.

  22. 23
    LimitsOfLanguage says:

    Jeffrey Gandee,

    Graduation rates have become targets, which potentially makes them subject to Goodhart’s law (“When a measure becomes a target, it ceases to be a good measure”).

    At Mizzou, there is a program to assist black men that is denied to black women and white people. That surely closes the gap between the graduation rates of black men and others, but that is by virtue of assistance given to them that is denied to others.

    However, at many other universities attempts to improve graduation rates seem to at least theoretically be neutral, helping all those in need. However, with the focus being on closing the racial gap, the incentive is to help black people more than white students who flounder. Often, when you set targets that can (only/best/also) be met by discrimination, then even if there is no official policy of discrimination, people will do what it takes to achieve the targets.

    Furthermore, if teachers have a common progressive belief, that black people have to work harder for the same outcome, even at the university, they may grade these students higher than the quality of their work merits.

    If more equal graduation rates are achieved by educating black people to a lower standard, this would then logically impact these people in their future careers, causing black graduates to do worse than white graduates (as the former are educated to a lower standard). Ironically, one can then expect this to be visible in the careers of black people, which could then be falsely be used as evidence for the claim of discrimination against black people, while the actual discrimination is against poorly educated graduates.

  23. 24
    lurker23 says:

    If you come from a worse school and/or a lower SES background, you had to work harder to get the same level of qualifications, and those qualifications should be judged accordingly.

    should they? that is true if the idea is to let people in because of “hard they work.” but i do not think that is how it is.

    when people are just kids then we care only about how hard they try. but when people are grownups we care mostly about what they can do, i do not care if my doctor tries, i care if they are good at being a doctor, same for most people. college is where things start to change from caring about trying to caring about doing, i think.

    of course if you look at someone who went to a bad school and got a bad education and does not know alot of the stuff school is supposed to teach then that person would probably know more if they went to a better school. but you do not know how MUCH more, i know alot of people in good schools and some of them do very very well and other people do not do well. i think the aa people just make up a number. (another fairer way is that you can try to use a test that will not so much look at what you know but instead will look at how smart you are and how well you think, so you might think “i will let smart good-thinking people in even if they have a bad education”. but the problem is that those tests are alot like iq tests. and those will not give the results that aa people want at all.)

    or, if people think think that the sat is really easy to learn, so you just give classes to alot of people and hope they will do well on the SAT. that makes the most sense because the sat class is only 100 or maybe 200 hours and the sat makes a big difference in admission. but that does not work well either, i think, because it turns out that the sat actually needs you to know things like math (you cannot learn trigonometry in 100 hours) and english (you cannot learn to read well or fast in 100 hours) and so maybe the whole “sat is just about expensive classes” thing is not true, either

    Ironically, one can then expect this to be visible in the careers of black people

    i think this is true. i read that we would have alot more black engineers and alot more black doctors with less aa because they keep letting people into supergood schools under aa where they are at the bottom of their class. but the people who graduate with an engineer degree are usually high up in the class because it is hard, so they drop out of that program and never become engineers.

    you can get an engineer degree at MIT and alot of other places, it is all the same certification. but it is ALOT harder to be in an mit class because 75% of MIT people have at least a 770 on the sat math section and at least 25% (probably more) have an 800.

    say you have an aa person who was good at high school math but did not get to take calculus in high school. and they are smart. and maybe they got a 1450 math score on that part of the sat. so they are very much in the top 1% of black students.

    and this person will do well at alot of places!!! but if they go to mit they STILL might be one of the worst students in the school and especially in the engineer program. so they will end up probably dropping engineering.

  24. 25
    desipis says:

    Harlequin:

    If you come from a worse school and/or a lower SES background, you had to work harder to get the same level of qualifications, and those qualifications should be judged accordingly. And we use race because we know simple measures of SES often don’t capture the full story of this background (e.g., how wealthy black kids tend to live in neighborhoods with high poverty rates even though they’re not poor themselves).

    Except we can directly the measure parent income, neighbourhoods and schools that students go to, so there’s no need to use race as a poor proxy for these things. Pretty much anything you can measure in a study to show it produces is something you can measure on all students and take into account when considering who to accept. The only reason to use race is if you don’t care about the underlying reality or injustices and instead just care about the numbers at a superficial level.

  25. 26
    Kate says:

    If I look at the original story (instead of the more biased website that you linked to), it’s argued that billed healthcare went down by $1.8 million. However, uncontracted healthcare is notoriously overbilled compared to actual cost (as noted by the very study that the story is partially based on). Presumably, very little of this is paid for by the homeless and instead, it is mostly written off and the costs of that care passed on to other ER care recipients. This results in overly high ER bills. If you reduce use of the ER by non-payers, you will actually save less than seems at first glance.

    It looks like either all the articles did a piss poor job of summarizing it, or there is more data we haven’t found yet. Despite your aggressive accusations, I am actually interested in knowing the truth about this. I’ll admit that I’ve been busy, and didn’t read the article as closely as I should have. Thanks for linking to the study.
    No one says billed healthcare went down by $1.8 million. I actually don’t know where the articles got the $1.8 million/year figure. According to the report, billed healthcare went down $2.4 million in the first two years (p.10). As you pointed out, they note that “Actual costs are typically less than the charges reflected in hospital billing data.” But, you fail to note costs NOT included in hospital billing data, including “additional amounts from physicians who bill for professional services separately” which they didn’t track (p. 10); and Medic utilization (p. 11), resulting in reduction of billed amount of $258,604. They also don’t calculate the savings on law enforcement, in reduced arrests (82%) and days spent in jail (89%, from 1180 to 130). This study reported the average daily cost per inmate in Mecklenburg County, NC as $166.04/day, a savings of $174,342.

    Presumably, very little of this is paid for by the homeless and instead, it is mostly written off and the costs of that care passed on to other ER care recipients.

    Actually, the study says that at the outset 36% of participants were covered by Medicaid (p.8), 2 were over 65, so covered by Medicare, and 9 were veterans, so covered by the VA (although there may be some overlap, p. 20). So, at least 1/3, and perhaps as many as half, weren’t uncontracted, non-payers at the outset.

    Also note that this particular facility does not in fact give people housing, but demands 30% of people’s income as rent.

    I think they were pretty transparent about that. I’m not sure why you see that as a problem.

    The total number of homeless in Charlotte seems to be around 600, but the article claims that only 200 qualify for this facility. So are they cherry picking the least problematic third?

    No, the reverse. They were cherry-picking the highest cost homeless, to try to produce the highest returns on investment. I think they were pretty transparent about that. Two quotes from the report:
    “Tenants were homeless an average of seven years prior to moving into Moore Place and experienced periods of homelessness ranging up to 25 years.” p.8
    “Participating tenants were particularly vulnerable regarding age, disabling conditions, and the impact of traumatic stress.” p. 59

    Also, there was a 15% dropout rate. These people went back to the streets.

    Given the initial population, I think that is a fantastic result.

  26. 27
    LimitsOfLanguage says:

    Kate,

    Despite your aggressive accusations, I am actually interested in knowing the truth about this.

    I’m not sure why you are saying this. I never accused you of being dishonest or not wanting to know the truth.

    I do think that you have a tendency to link to very progressive sites whose biases are in line with the thing the link is supposed to show. They therefor have a tendency to be biased to present the evidence in the best possible light for their (and your) argument, sometimes even to such an extent that it can be called a falsehood.

    My comment was intended to make you consider looking for sources that are a bit less biased in the direction of the argument that is being made. I try to do the same thing with my sources when criticizing progressive excesses, often linking to sites like WaPo and the NYT. This is not a coincidence. I intentionally strongly favor sites whose biases go against my argument. Their presentation of the evidence in favor of my argument is often going to be robust, since they are not going to write down claims in favor of my argument unless the evidence for it is really strong.

    No one says billed healthcare went down by $1.8 million. I actually don’t know where the articles got the $1.8 million/year figure. According to the report, billed healthcare went down $2.4 million in the first two years (p.10).

    The ThinkProgress article you linked to, itself links to a Charlotte Observer article as its source of information, which gives this figure:

    The study, conducted by the university’s Department of Social Work, found Moore Place saved $1.8 million in its first year by drastically reducing the amount of time its tenants spent in emergency rooms (447 fewer visits) and admitted to hospitals (372 fewer days).

    Page 10 is a summary of the sub-reports that are reproduced later on. The $1.8 million used in the article seems to come from table 24 on page 37, where one of the two hospital systems saved $1.8m in billed care. I don’t understand why the article picked this figure though, because I don’t see a significant difference between the two hospital systems. The most likely explanation seems to be a hasty journalist making an error. As an aside, this is why I often look for original sources, because mistakes, cherry picking or other such issues are quite common.

    But, you fail to note costs NOT included in hospital billing data, including […] They also don’t calculate the savings on law enforcement

    I actually addressed this last bit in my comment: “Of course, the article also didn’t quantify the costs in reducing policing costs.”

    My argument is not that the costs are necessarily higher than the the savings, but that the presented evidence is too poor to draw conclusions one way or the other. Claims that it is certain that the savings are higher or even 3 times higher are thus not justified.

    Ultimately, I think that people are making a bad choice when making such unjustified claims that try to present policies as a slam-dunk with no real downsides. It polarizes society. On the one hand you have people who already strongly favor the policies who become convinced that no proper counterarguments exists and that everyone on the other side is thus completely stupid; while on the other hand you have people who are more critical of the policies who get frustrated that the response to their concerns consists of unjustified claims based on falsehoods and who can become convinced that those who favor the policies don’t care about facts.

    I think they were pretty transparent about [asking for rent]. I’m not sure why you see that as a problem.

    We started off with a claim that giving people housing was a good thing to do. When I pushed back, you gave me this example where the housing is not given, but very heavily subsidized. This is less progressive and more in line with a bit more conservative/moderate thinking, where a quid-pro-quo tends to be more favored (and assumed to work better).

    In general, it seems like certain framing works better to convince people with certain ideologies. The reason why I pointed out the rent was actually not so much for your benefit, but to try to make certain other, more conservative/moderate readers more sympathetic to the idea of helping the homeless with housing.

    No, the reverse. They were cherry-picking the highest cost homeless, to try to produce the highest returns on investment. I think they were pretty transparent about that. Two quotes from the report:

    Good catch, I missed that. Although…I actually said least problematic, not least costly. People tend to age out of crime and such, so it’s still possible that the solution works for the selected group, but not so well for other homeless people. However, the quotes do strongly suggest that the people were not selected for having few problems, which makes this far less likely.

    Given the initial population, I think that [a 15% dropout rate] is a fantastic result.

    Sure, but it would still mean that if the solution is implemented for all ~600 homeless people in Charlotte and works equally well for them, you have 90 people for whom it doesn’t work. For all of America, that would be 83,000 people still on the streets.

    Of course, solving 85% of a problem is great, but it still means we need other solutions for the remaining 15%.

  27. 28
    LimitsOfLanguage says:

    lurker23,

    i read that we would have alot more black engineers and alot more black doctors with less aa because they keep letting people into supergood schools under aa where they are at the bottom of their class. but the people who graduate with an engineer degree are usually high up in the class because it is hard, so they drop out of that program and never become engineers.

    When it comes to role models & such, it is a good question what is better: to have some more black people with an Ivy degree who are often poor to mediocre in their field at a cost of many people dropping out or way more good black engineers & doctors with a less prestigious degree.

    Interestingly, black people who do well seem to be far less eager than white people to escape low income neighborhoods, which should theoretically mean that black people who do well make for better role models in their neighborhoods than white people who do well (unless role models don’t work that way*).

    * Research into role models tends to find that the impact is not very impressive.

    but if they go to mit they STILL might be one of the worst students in the school and especially in the engineer program. so they will end up probably dropping engineering.

    Indeed. It may be better for black people if more of them become engineers & doctors at second-tier universities, rather than have these people get less practical Ivy diplomas, while their spots at the second-tier universities are taken by others who are also below average for that environment and who also tend to pick easy majors. Engineering jobs are more easily to be found even for those with few connections, so this could help a lot with upward mobility. I suspect that (less qualified) black people who mostly are from a less advantaged background get less of a benefit from the networking opportunities at Ivy league universities anyway, just like others from less advantaged backgrounds.

    In the West many seem to believe that the barriers between the classes are overwhelmingly access-related, but my experience and observation is that culture and wealth-based barriers are very significant. IMO, more attention should be paid to these, rather than to assume that merely getting people access to a path of success for the better off, will automatically allow these people to travel that same path.

  28. 29
    Nancy Lebovitz says:

    I’ve also seen horror stories about remedial education for college students where there’s a failure to check on what the student already knows. This is dispiriting for students who actually want to learn things and don’t have time to waste.

    I can easily believe both that students go in ill-prepared and that remedial education is sloppily executed.

  29. 30
    Ampersand says:

    The only reason to use race is if you don’t care about the underlying reality or injustices and instead just care about the numbers at a superficial level.

    Or if you care about racism. (In addition to, not instead of, economic class.)

  30. 31
    desipis says:

    Or if you care about racism. (In addition to, not instead of, economic class.)

    What form of racism are you talking about? If you’re talking about direct explicit racism, then sure that should be tackled head on (albeit in a racially neutral way, such as simply prohibiting racial discrimination).

    However, my comment was addressing Harlequin’s observation about SES status and going to worse schools. Addressing those issues directly will indirectly address the indirect forms of racism, without needing to directly discriminate based on race.

Leave a Reply

Your email address will not be published. Required fields are marked *