home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky alt.discrimination:5663 soc.culture.african.american:14198
- Newsgroups: alt.discrimination,soc.culture.african.american
- Path: sparky!uunet!haven.umd.edu!purdue!mentor.cc.purdue.edu!pop.stat.purdue.edu!hrubin
- From: hrubin@pop.stat.purdue.edu (Herman Rubin)
- Subject: Re: Institutional racism
- Message-ID: <C17zF8.HzM@mentor.cc.purdue.edu>
- Sender: news@mentor.cc.purdue.edu (USENET News)
- Organization: Purdue University Statistics Department
- References: <1993Jan20.021235.22531@samba.oit.unc.edu> <ISBELL.93Jan20113253@panther.ai.mit.edu>
- Distribution: usa
- Date: Thu, 21 Jan 1993 19:57:05 GMT
- Lines: 55
-
- In article <ISBELL.93Jan20113253@panther.ai.mit.edu> isbell@ai.mit.edu (Charles L Isbell) writes:
- >Terry.Parks@launchpad.unc.edu (Terry Parks) writes:
- >|A claim was made of the existence of a mortage lending study which shows
- >|a racial bias (actually an alleged racial bias in favor of Caucasians)
- >|which didn't cook data. However, the study, reported in the October 9th, 1992
- >|Boston Globe does just that.
-
- >| Study shows racial bias in lending
-
- >| A landmark study released yesterday shows that banks in Greater Boston
- >| discriminate against black and Hispanic mortage applicants, offering
- >| the most damning evidence to date of racial hurdles facing
- >| minority homebuyers.
-
- >|But what is this, in a paragraph back on page 5:
-
- >| Syron and Munnell [the report presenters] stressed that the new
- >| study found no evidence that banks discriminate against minority
- >| mortgage applicants with "perfect" financial and employment
- >| records.
-
- >And exactly *how* is this "cooking" the data?
-
- >And exactly *how* does this mean that there is "no discrimination"?
-
- >|So there is no discrimination, so why the big headline? Apparently the
- >|discrimination is that a person with twenty late credit card payments
- >|was more often denied than a person with just one. Apply a flame and
- >|voila, cooked data. Also part of the creative baking was that property
- >|location was totally ignored. Is it plausible that a $100,000 house in
- >|a run-down block of $50,000 houses would be more of a mortgage risk than
- >|a $100,000 house in a block of $200,000 houses?
-
- >How utterly and completely pathetic. When folks have "perfect"
- >records, banks hop all over them. When folks have a non-perfect
- >record (>80% of the applicants, btw), Whites are more likely than
- >Blacks to be approved.
-
- >The article makes perfectly clear that only groups of equal applicants
- >were compared.
-
- How do you decide that applicants are equal? There is no way to get a
- complete list of all of the criteria involved; we have no way to codify
- the information into an intelligible form. I have looked at far too
- many applicants for support for graduate school not to realize how
- inadequate the information available is for the purpose.
-
- The information supplied is not the person. I suggest that if you have
- the necessary background, you look at a sound statistics book, and look
- at the notion of "regression." It is valid here.
- --
- Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
- Phone: (317)494-6054
- hrubin@snap.stat.purdue.edu (Internet, bitnet)
- {purdue,pur-ee}!snap.stat!hrubin(UUCP)
-