Jump to content

For the experts on the board


Catdaddyhorn

Recommended Posts

3 minutes ago, Parliament said:

Why are you CR'ing Bada Bing?

What exactly does this have to do with politics? I created the shaggy thread on the history of American housing policy in Bada Bing therefore I'm posting this one here.  The very idea that anytime someone post an article addressing black Americans it's immediately considered political or controversial in some manner is telling.  Would you ask the same question if I started a thread here on the opioid crisis in middle America? If not ask yourself why? 

  • Like 2
Link to comment
Share on other sites

21 minutes ago, Incredulity said:

It may well be true, but I certainly can’t find Jacob William Faber’s research associated with that specific statistic.  Would be awful nice if a link or some reference was provided.

https://www.tandfonline.com/doi/abs/10.1080/10511482.2013.771788

I'm still looking, but scanning through the abstract this may be one of the references 

Link to comment
Share on other sites

We aren't talking about a single black family here? We're talking about a large sample of black families being more likely to be thrown into subprime pools than white families making 170k less than them.  No amount of bending over backwards to stick your head in the sand to avoid seeing racism can hide this fact.  Not to mention what we already know in regards to this topic

 

Then throw in the fact that this country has only attempted to not be racist as fuck for about 50 years of its history.  So yeah I'm going take the large stockpile of evidence and say racism has something to do with it. 

Edited by Catdaddyhorn
Link to comment
Share on other sites

What's the point of this thread?  You're asking people's opinions if this is institutional racism and when people say "no", you're arguing with them that they're wrong.  Just tell me what I'm supposed to say so you will STFU.

  • Like 2
Link to comment
Share on other sites

1 hour ago, Assman said:

What's the point of this thread?  You're asking people's opinions if this is institutional racism and when people say "no", you're arguing with them that they're wrong.  Just tell me what I'm supposed to say so you will STFU.

Just think of this as another post within the thread on the old site. Continuing the conversation so to speak. 

http://www.shaggytexas.com/board/showthread.php/179289-AmericaHousing-Policies-impact-on-the-Black-Community?p=10934987&viewfull=1#post10934987

Link to comment
Share on other sites

2 hours ago, Catdaddyhorn said:

What exactly does this have to do with politics? I created the shaggy thread on the history of American housing policy in Bada Bing therefore I'm posting this one here.  The very idea that anytime someone post an article addressing black Americans it's immediately considered political or controversial in some manner is telling.  Would you ask the same question if I started a thread here on the opioid crisis in middle America? If not ask yourself why? 

Both subjects have very strong drivers within the political realm.  It would appear pretty difficult to have a thorough discussion of either topic without mentioning said political influences.

 

Good luck.  We're counting on you.

Link to comment
Share on other sites

10 minutes ago, NotActuallyALonghorn said:

Doe having a low credit score affect one's ability to make an intelligible thread title that gives an indication about what said thread may be about?

Surly expert with a straight face: "Well your 200k families have lower credit scores than the 30k families"

catdaddyhorn: "That's not the case. In fact, banking data shows over decades this isn't the case." Black families with higher credit scores and better income consistently get less favorable loans that white families" 

Another Surly expert with an even more sober and earnest face: "Well all your 200k families have lower credit scores than the 30k families"

Link to comment
Share on other sites

Here's something more directly on point, although not a "nationwide," thing.  https://prospect.org/article/staggering-loss-black-wealth-due-subprime-scandal-continues-unabated

Unabashedly racist lending practices by WF.

Also notes a hugely high mortgage default rate in PG county, which isn't entirely explained by subprime loans.

Link to comment
Share on other sites

It’s not racism.

I got a loan to buy a house in 2004 from Countrywide Home Loans, then largest home mortgage lender in the county. I did the whole thing by phone/fax/mail. I doubt they knew if I was black or white until long after the loan terms were settled. 

I will add this though...

Regardless of race, there is an amazing amount of ignorance about how the whole process of creating a home mortgage works. A huge majority of borrowers have no idea where the money comes from; who the players are, or how the players get compensated. Very few are aware that the friendly person on the phone does NOT have your best interest at heart. The higher your interest rate, they more the lender profits when the loan gets sold, and they all get sold. Every borrower should shop around for a loan, but few do. Lenders take advantage of the ignorant. When you sit down to buy a car, you know the dealer’s financial motivations are diametrically opposed to your own. It’s the exact same thing when dealing with a home mortgage originator, but most folks don’t know it.  

Bernard

  • Like 1
Link to comment
Share on other sites

3 hours ago, Catdaddyhorn said:

We aren't talking about a single black family here? We're talking about a large sample of black families being more likely to be thrown into subprime pools than white families making 170k less than them.  No amount of bending over backwards to stick your head in the sand to avoid seeing racism can hide this fact.  Not to mention what we already know in regards to this topic

You don’t seem to grasp that the subprime mortgages were not about income. They were about credit scores. The article talks about income, but not credit scores. The research behind the article says they examined four million loan applications. Where is the credit score data? It’s conveniently left out. If the data showed that black borrowers were steered into subprime loans while white borrowers with the same credit scores were considered prime, then we’d have a real, live smoking gun. I can only assume that data, since it was left out of the article, doesn’t exist. 

Bernard

Edited by Bernard
Link to comment
Share on other sites

The articles pretty uniformly indicate that, adjusted for credit scores and other neutral factors, "minorities" received "subprime" loans at a higher rate than whiteys.

Quote

The Center for Responsible Lending found that during the housing boom, 6.2 percent of whites with a credit score of 660 and higher received high-interest mortgages but 21.4 percent of blacks with a score of 660 or higher received these same loans.

It turned out that several of the major banks had been purposely giving people of color subprime mortgages, including borrowers who would have qualified for a prime loan. The City of Baltimore took Wells Fargo to court, bringing some of the banking giant’s abhorrent lending practices to light. One former employee testified that in 2001, Wells Fargo created a unit that would be responsible for pushing expensive refinance loans on black customers, especially those living in Baltimore, southeast Washington, D.C., and Prince George’s County—all locations with large black populations. 

According to court testimony, some of the loan officers at Wells Fargo spoke of these subprime loans as “ghetto loans,” and referred to their black customers as “mud people.” There was even a cash incentive for loan officers to aggressively market subprime mortgages in minority neighborhoods. In the end, the Justice Department found that 4,500 homeowners in Baltimore and the Washington, D.C., region that had been affected by these flat-out racist lending practices.

Specifically targeted for subprime loans among the minority demographic were black women. Women of color are the most likely to receive subprime loans while white men are the least likely; the disparity grows with income levels. Compared to white men earning the same level of income, black women earning less than the area median income are two and a half times more likely to receive subprime. Upper-income black women were nearly five times more likely to receive subprime purchase mortgages than upper-income white men.

 

Edited by TwiceHorn
  • Like 2
Link to comment
Share on other sites

45 minutes ago, Bernard said:

It’s not racism.

I got a loan to buy a house in 2004 from Countrywide Home Loans, then largest home mortgage lender in the county. I did the whole thing by phone/fax/mail. I doubt they knew if I was black or white until long after the loan terms were settled. 

I will add this though...

Regardless of race, there is an amazing amount of ignorance about how the whole process of creating a home mortgage works. A huge majority of borrowers have no idea where the money comes from; who the players are, or how the players get compensated. Very few are aware that the friendly person on the phone does NOT have your best interest at heart. The higher your interest rate, they more the lender profits when the loan gets sold, and they all get sold. Every borrower should shop around for a loan, but few do. Lenders take advantage of the ignorant. When you sit down to buy a car, you know the dealer’s financial motivations are diametrically opposed to your own. It’s the exact same thing when dealing with a home mortgage originator, but most folks don’t know it.  

Bernard

Ignore the peer reviewed, academic data saying otherwise. This guy got a loan over the phone. 

 

  • Like 1
  • Haha 1
Link to comment
Share on other sites

3 hours ago, Incredulity said:

Well if you find the source data please post it.  I ain’t paying $43 to download that paper.

Here's the description of the data set from the Faber article.

Spoiler

The HMDA

The HMDA was passed in 1975 to shed light on the mortgage industry. The law requires the large majority (Avery et al., 2007) of lending institutions to report information regarding home loan applications on a yearly basis (Pettit & Droesch, 2008). The HMDA covered 8,886 lending institutions in 2006 (Avery et al., 2007). This rich data set provided by the Federal Financial Institutions Examination Council (FFIEC, 2007a, 2007b) includes borrower characteristics (e.g., gender, race, ethnicity, income), information about the home (e.g., census tract, whether it is owner-occupied, property type), loan decision details (e.g., whether the loan was offered, reason for denial, rate spread, loan amount), and census tract –level data (e.g., percentage of the tract population that is of a minority racial group, median family income, population). A few variables are absent from the HMDA data set, including loan-to-value ratio, applicant credit score, and down payment size, which are typically important when a lender is assessing the riskiness of a potential borrower (Avery et al., 2007; Bocian, Ernst, & Li, 2007). Although this means my Housing Policy Debate 333 estimates are imperfect, HMDA data are the strongest data available because they are the only public, national mortgage data that include borrower race and application neighborhood (Bradford, 2002).2

Beginning in 1989, the HMDA began requiring lenders to ask applicants their race and ethnicity; however, applicants could refuse to answer (Ross & Yinger, 2002). If an applicant refused to answer, the lender could attempt a guess, based on physical characteristics or name. Unfortunately, there is no indication of which party was responsible for the race data. Earlier analyses were hampered by the fact that lenders were not required to request this information for applications processed via telephone, mail, or electronic means, which constituted a growing proportion of mortgage applications through the 1990s (Dietrich, 2002; Huck, 2001). Analysis showed these missing data to be more of a concern in studies of refinance loans than in the studies of purchase loans (Huck, 2001). A 2003 change in reporting requirements forced lenders to request these data from all applicants. The proportion of home-purchase applications missing race or ethnicity information fell from about 18% in 2001 to 12% in 2003 (FFIEC, 2004).3 I coded all respondents into six categories, based on ethnicity and the first race identified.4

Because I was primarily interested in the role that credit plays in expanding homeownership opportunities, I focused on home-purchase loans. This was in contrast to much of the previous research on subprime lending, which has concentrated on refinance loans, but in line with aforementioned concerns regarding housing credit access from earlier eras, which typically focused on home-purchase lending (Scheessele, 2002). I also limited models to owner-occupied borrowers because it requires fewer assumptions about motivations and the impact of subprime loans. Finally, this sampling decision reduced missing race data problems, which have been more common among refinancing loans (Huck, 2001).

Within home-purchase loans, I excluded second liens. This sampling decision was made because it is otherwise impossible to avoid double-counting individuals applying for piggy-back loans, which may have been taken out to avoid mortgage insurance on low down payment loans or to stay within the conforming loan limit (Avery et al., 2007; Been et al., 2008; Pettit & Droesch, 2008). This decision also meant that the loan amounts reported where there were second mortgages did not measure the full credit risk taken by the borrower. It is possible that for those applicants who applied simultaneously for first and second lien loans, the relationship between loan amount and denial or subprime likelihood would have appeared weaker in statistical models because their full risk profile would not have been captured. If this were the case, the statistical consequence of excluding second liens should have biased any effect of loan amount on denial or subprime approval toward 0. This certainly suggests caution when interpreting the relationship between loan amount and application outcomes.

I only included applications for owner-occupied borrowers because loans for other properties were likely occupied by renters or acted as second homes, which is inconsistent with my concern about homeownership access (Avery et al., 2007; Kingsley & Pettit, 2009). Finally, because conventional loans were the large majority of home-purchase loans and, historically, concerns regarding lender fairness have focused on the conventional market (Canner et al., 1999), I chose to leave out Federal Housing Administration, Veterans Administration, Farm Service Agency, and Rural Housing Service loans. Subprime representation was lower among government-backed loans, but because they made up such a small percentage (only 5%) of loan activity in 2006 (Avery et al., 2007), their exclusion from my sample should not have substantively biased my findings in the same way as it may have in analyses of earlier years.5 334 J.W. Faber

My final data set consisted of 3,819,923 loan applications, of which 1,590,267 (41.63%) were denied, 2,023,247 (52.97%) were approved with a prime rate, and 206,409 (5.40%) were approved with a subprime rate. A loan was defined as subprime if it had an interest rate 3 or more points above the federal treasury rate. This definition was consistent with the extant literature, which often used the term high-cost loan synonymously (Been et al., 2008; Kingsley & Pettit, 2009).

The data in Figure 1 present the basic relationship explored by this article. Subprime lending rates varied substantially by race, with Latino borrowers (7.52%) having seen a rate nearly 3 times that of Asian borrowers (2.57%). Mortgage applications by whites (5.10%) resulted in a subprime loan less often than the overall average (5.40%), whereas applications by blacks (7.04%) were above average. Table 1 displays means for each of the covariates included in the regression models employed for the whole sample and for all the applications that resulted in subprime approval. Not surprisingly, there were many notable differences. Black (14.2%) and Latino (20.3%) representations are much higher among subprime originations than in the full sample, whereas those of whites (53.5%) and Asians (2.4%) were lower. Applications with missing race data were less common among subprime originations. Both applicant ($81,388) and neighborhood ($59,623) incomes were lower among subprime origination as well. The average amount borrowed ($190,891) was much lower among subprime applicants. The tracts in which subprime loans originated were less white (69.4%) than in the full sample.


Here is the main result table and its description.

Spoiler

 

  Model 1 Model 2 Model 3 Model 4
Outcome: Loan denial
  Blacka 3.587 (0.022)*** 2.797 (0.017)*** 1.757 (0.058)*** 3.086 (0.030)***
  Latino 2.875 (0.017)*** 2.022 (0.012)*** 0.683 (0.024)*** 2.275 (0.021)***
  Asian 1.297 (0.011)*** 1.067 (0.009)*** 0.258 (0.012)*** 1.145 (0.016)***
  Other 2.148 (0.041)*** 1.670 (0.031)*** 1.701 (0.032)*** 1.619 (0.030)***
  No race identified 2.424 (0.024)*** 2.223 (0.023)*** 2.226 (0.023)*** 2.186 (0.022)***
  Multiracial 0.937 (0.021)** 0.946 (0.021)* 0.950 (0.021)* 0.939 (0.021)**
  Female 1.127 (0.003)*** 1.021 (0.003)*** 1.016 (0.003)*** 1.021 (0.003)***
  Log (income) 0.913 (0.003)*** 0.940 (0.003)*** 0.883 (0.004)*** 0.940 (0.003)***
  Has co-applicant   0.618 (0.002)*** 0.622 (0.002)*** 0.620 (0.002)***
  Log (loan amount)   1.151 (0.005)*** 1.153 (0.005)*** 1.149 (0.005)***
  Tract percentage of    people of color   1.004 (0.000)*** 1.004 (0.000)*** 1.006 (0.000)***
  Log (tract income)   0.592 (0.006)*** 0.592 (0.006)*** 0.591 (0.006)***
  In MSA/MD   0.840 (0.007)*** 0.851 (0.007)*** 0.835 (0.007)***
  Mid-Atlantic   1.074 (0.012)*** 1.064 (0.012)*** 1.070 (0.012)***
  East North Central   1.051 (0.011)*** 1.041 (0.011)*** 1.050 (0.011)***
  West North Central   0.954 (0.013)*** 0.943 (0.013)*** 0.955 (0.013)***
  South Atlantic   0.981 (0.011) 0.974 (0.011)* 0.968 (0.011)**
  East South Central   1.060 (0.014)*** 1.047 (0.014)*** 1.054 (0.014)***
  West South Central   1.147 (0.014)*** 1.160 (0.014)*** 1.132 (0.014)***
  Mountain   1.243 (0.018)*** 1.249 (0.018)*** 1.227 (0.018)***
  Pacific   1.369 (0.016)*** 1.335 (0.016)*** 1.355 (0.016)***
  Black × log income     1.115 (0.009)***  
  Latino × log income     1.285 (0.010)***  
  Asian × log income     1.366 (0.014)***  
  Black × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Latino × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Asian × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Constant 0.733 (0.011)*** 130.792 (13.789)*** 167.191 (17.397)*** 130.451 (13.849)***
Base outcome: Prime approval        
Outcome: Subprime approval        
  Black 2.559 (0.022)*** 2.415 (0.023)*** 0.731 (0.038)*** 2.576 (0.039)***
  Latino 2.601 (0.022)*** 2.450 (0.022)*** 0.966 (0.051) 2.457 (0.036)***
  Asian 0.624 (0.011)*** 0.723 (0.012)*** 0.067 (0.008)*** 0.615 (0.018)***
  Other 1.462 (0.057)*** 1.609 (0.063)*** 1.648 (0.065)*** 1.614 (0.063)***
  No race identified 1.191 (0.021)*** 1.165 (0.021)*** 1.165 (0.021)*** 1.165 (0.021)***
  Multiracial 0.875 (0.042)** 0.855 (0.042)** 0.858 (0.042)** 0.853 (0.041)**
  Female 0.944 (0.005)*** 0.920 (0.005)*** 0.917 (0.005)*** 0.920 (0.005)***
  Log (income) 0.625 (0.003)*** 1.024 (0.006)*** 0.941 (0.006)*** 1.024 (0.006)***
  Has co-applicant   0.651 (0.004)*** 0.658 (0.004)*** 0.651 (0.004)***
  Log (loan amount)   0.780 (0.005)*** 0.780 (0.005)*** 0.779 (0.005)***
  Tract percentage of people    of color   0.998 (0.000)*** 0.999 (0.000)*** 0.998 (0.000)***
  Log (tract income)   0.412 (0.006)*** 0.413 (0.006)*** 0.413 (0.006)***
  In MSA/MD   0.681 (0.008)*** 0.690 (0.008)*** 0.681 (0.008)***
  Mid-Atlantic   1.027 (0.022) 1.013 (0.021) 1.024 (0.022)
  East North Central   1.357 (0.026)*** 1.343 (0.026)*** 1.357 (0.026)***
  West North Central   1.481 (0.033)*** 1.465 (0.032)*** 1.480 (0.033)***
  South Atlantic   1.176 (0.023)*** 1.167 (0.022)*** 1.173 (0.023)***
  East South Central   1.546 (0.035)*** 1.540 (0.035)*** 1.543 (0.035)***
  West South Central   1.787 (0.037)*** 1.793 (0.037)*** 1.779 (0.037)***
  Mountain   1.113 (0.026)*** 1.113 (0.026)*** 1.112 (0.026)***
  Pacific   1.286 (0.027)*** 1.253 (0.026)*** 1.277 (0.027)***
  Black × log income     1.337 (0.017)***  
  Latino × log income     1.243 (0.015)***  
  Asian × log income     1.699 (0.044)***  
  Black × neighborhood's    percentage of nonwhites       0.999 (0.000)***
  Latino × neighborhood's    percentage of nonwhites       1.000 (0.000)
  Asian × neighborhood's    percentage of nonwhites       1.004 (0.001)
  Constant 0.607 (0.014)*** 6,201.882 (909.609)*** 8,493.775 (1,238.971)*** 6,120.317 (899.170)***
Observations 3,341,693 3,340,428 3,340,428 3,340,428
Pseudo R 2 .044 .066 .067 .066

 

Description: 

Table 2 presents results from the four multinomial logits where prime approval is the base outcome. The top panel compares the outcome of loan denial with prime approval, and the bottom panel compares subprime approval with prime approval. Because the findings are odds ratios, each coefficient should be interpreted as the effect a one-unit change in the covariate has on the odds of the corresponding outcome being the result as opposed to the outcome of prime mortgage approval.88. If a coefficient is greater than 1, an increase in the variable increases the odds. If a coefficient is less than 1, an increase in the variable decreases the odds.View all notes 

For example, comparing black applicants to identical white applicants with respect to the measures controlled for in model 1, outcomes of their applications differed substantially: Blacks were 3.6 times more likely than whites to be denied a loan, and if approved, they were 2.6 times more likely to be offered a subprime loan rather than a prime loan. The first part of this section will discuss the basic, noninteracted models (models 1 and 2). The second part will consider the borrower race interactions with income and neighborhood racial composition (models 3 and 4). This section focuses on the primary variables of interest: borrower race, borrower income, neighborhood racial makeup, and interaction terms of race with income and race with neighborhood racial makeup.

 

 

Edited by deac_tracy
  • Like 1
Link to comment
Share on other sites

10 minutes ago, deac_tracy said:

Here's the description of the data set from the Faber article.

  Reveal hidden contents

The HMDA

The HMDA was passed in 1975 to shed light on the mortgage industry. The law requires the large majority (Avery et al., 2007) of lending institutions to report information regarding home loan applications on a yearly basis (Pettit & Droesch, 2008). The HMDA covered 8,886 lending institutions in 2006 (Avery et al., 2007). This rich data set provided by the Federal Financial Institutions Examination Council (FFIEC, 2007a, 2007b) includes borrower characteristics (e.g., gender, race, ethnicity, income), information about the home (e.g., census tract, whether it is owner-occupied, property type), loan decision details (e.g., whether the loan was offered, reason for denial, rate spread, loan amount), and census tract –level data (e.g., percentage of the tract population that is of a minority racial group, median family income, population). A few variables are absent from the HMDA data set, including loan-to-value ratio, applicant credit score, and down payment size, which are typically important when a lender is assessing the riskiness of a potential borrower (Avery et al., 2007; Bocian, Ernst, & Li, 2007). Although this means my Housing Policy Debate 333 estimates are imperfect, HMDA data are the strongest data available because they are the only public, national mortgage data that include borrower race and application neighborhood (Bradford, 2002).2

Beginning in 1989, the HMDA began requiring lenders to ask applicants their race and ethnicity; however, applicants could refuse to answer (Ross & Yinger, 2002). If an applicant refused to answer, the lender could attempt a guess, based on physical characteristics or name. Unfortunately, there is no indication of which party was responsible for the race data. Earlier analyses were hampered by the fact that lenders were not required to request this information for applications processed via telephone, mail, or electronic means, which constituted a growing proportion of mortgage applications through the 1990s (Dietrich, 2002; Huck, 2001). Analysis showed these missing data to be more of a concern in studies of refinance loans than in the studies of purchase loans (Huck, 2001). A 2003 change in reporting requirements forced lenders to request these data from all applicants. The proportion of home-purchase applications missing race or ethnicity information fell from about 18% in 2001 to 12% in 2003 (FFIEC, 2004).3 I coded all respondents into six categories, based on ethnicity and the first race identified.4

Because I was primarily interested in the role that credit plays in expanding homeownership opportunities, I focused on home-purchase loans. This was in contrast to much of the previous research on subprime lending, which has concentrated on refinance loans, but in line with aforementioned concerns regarding housing credit access from earlier eras, which typically focused on home-purchase lending (Scheessele, 2002). I also limited models to owner-occupied borrowers because it requires fewer assumptions about motivations and the impact of subprime loans. Finally, this sampling decision reduced missing race data problems, which have been more common among refinancing loans (Huck, 2001).

Within home-purchase loans, I excluded second liens. This sampling decision was made because it is otherwise impossible to avoid double-counting individuals applying for piggy-back loans, which may have been taken out to avoid mortgage insurance on low down payment loans or to stay within the conforming loan limit (Avery et al., 2007; Been et al., 2008; Pettit & Droesch, 2008). This decision also meant that the loan amounts reported where there were second mortgages did not measure the full credit risk taken by the borrower. It is possible that for those applicants who applied simultaneously for first and second lien loans, the relationship between loan amount and denial or subprime likelihood would have appeared weaker in statistical models because their full risk profile would not have been captured. If this were the case, the statistical consequence of excluding second liens should have biased any effect of loan amount on denial or subprime approval toward 0. This certainly suggests caution when interpreting the relationship between loan amount and application outcomes.

I only included applications for owner-occupied borrowers because loans for other properties were likely occupied by renters or acted as second homes, which is inconsistent with my concern about homeownership access (Avery et al., 2007; Kingsley & Pettit, 2009). Finally, because conventional loans were the large majority of home-purchase loans and, historically, concerns regarding lender fairness have focused on the conventional market (Canner et al., 1999), I chose to leave out Federal Housing Administration, Veterans Administration, Farm Service Agency, and Rural Housing Service loans. Subprime representation was lower among government-backed loans, but because they made up such a small percentage (only 5%) of loan activity in 2006 (Avery et al., 2007), their exclusion from my sample should not have substantively biased my findings in the same way as it may have in analyses of earlier years.5 334 J.W. Faber

My final data set consisted of 3,819,923 loan applications, of which 1,590,267 (41.63%) were denied, 2,023,247 (52.97%) were approved with a prime rate, and 206,409 (5.40%) were approved with a subprime rate. A loan was defined as subprime if it had an interest rate 3 or more points above the federal treasury rate. This definition was consistent with the extant literature, which often used the term high-cost loan synonymously (Been et al., 2008; Kingsley & Pettit, 2009).

The data in Figure 1 present the basic relationship explored by this article. Subprime lending rates varied substantially by race, with Latino borrowers (7.52%) having seen a rate nearly 3 times that of Asian borrowers (2.57%). Mortgage applications by whites (5.10%) resulted in a subprime loan less often than the overall average (5.40%), whereas applications by blacks (7.04%) were above average. Table 1 displays means for each of the covariates included in the regression models employed for the whole sample and for all the applications that resulted in subprime approval. Not surprisingly, there were many notable differences. Black (14.2%) and Latino (20.3%) representations are much higher among subprime originations than in the full sample, whereas those of whites (53.5%) and Asians (2.4%) were lower. Applications with missing race data were less common among subprime originations. Both applicant ($81,388) and neighborhood ($59,623) incomes were lower among subprime origination as well. The average amount borrowed ($190,891) was much lower among subprime applicants. The tracts in which subprime loans originated were less white (69.4%) than in the full sample.


Here's the main result table and the description:

  Reveal hidden contents

 

  Model 1 Model 2 Model 3 Model 4
Outcome: Loan denial
  Blacka 3.587 (0.022)*** 2.797 (0.017)*** 1.757 (0.058)*** 3.086 (0.030)***
  Latino 2.875 (0.017)*** 2.022 (0.012)*** 0.683 (0.024)*** 2.275 (0.021)***
  Asian 1.297 (0.011)*** 1.067 (0.009)*** 0.258 (0.012)*** 1.145 (0.016)***
  Other 2.148 (0.041)*** 1.670 (0.031)*** 1.701 (0.032)*** 1.619 (0.030)***
  No race identified 2.424 (0.024)*** 2.223 (0.023)*** 2.226 (0.023)*** 2.186 (0.022)***
  Multiracial 0.937 (0.021)** 0.946 (0.021)* 0.950 (0.021)* 0.939 (0.021)**
  Female 1.127 (0.003)*** 1.021 (0.003)*** 1.016 (0.003)*** 1.021 (0.003)***
  Log (income) 0.913 (0.003)*** 0.940 (0.003)*** 0.883 (0.004)*** 0.940 (0.003)***
  Has co-applicant   0.618 (0.002)*** 0.622 (0.002)*** 0.620 (0.002)***
  Log (loan amount)   1.151 (0.005)*** 1.153 (0.005)*** 1.149 (0.005)***
  Tract percentage of    people of color   1.004 (0.000)*** 1.004 (0.000)*** 1.006 (0.000)***
  Log (tract income)   0.592 (0.006)*** 0.592 (0.006)*** 0.591 (0.006)***
  In MSA/MD   0.840 (0.007)*** 0.851 (0.007)*** 0.835 (0.007)***
  Mid-Atlantic   1.074 (0.012)*** 1.064 (0.012)*** 1.070 (0.012)***
  East North Central   1.051 (0.011)*** 1.041 (0.011)*** 1.050 (0.011)***
  West North Central   0.954 (0.013)*** 0.943 (0.013)*** 0.955 (0.013)***
  South Atlantic   0.981 (0.011) 0.974 (0.011)* 0.968 (0.011)**
  East South Central   1.060 (0.014)*** 1.047 (0.014)*** 1.054 (0.014)***
  West South Central   1.147 (0.014)*** 1.160 (0.014)*** 1.132 (0.014)***
  Mountain   1.243 (0.018)*** 1.249 (0.018)*** 1.227 (0.018)***
  Pacific   1.369 (0.016)*** 1.335 (0.016)*** 1.355 (0.016)***
  Black × log income     1.115 (0.009)***  
  Latino × log income     1.285 (0.010)***  
  Asian × log income     1.366 (0.014)***  
  Black × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Latino × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Asian × neighborhood's    percentage of nonwhites       0.997 (0.000)***
  Constant 0.733 (0.011)*** 130.792 (13.789)*** 167.191 (17.397)*** 130.451 (13.849)***
Base outcome: Prime approval        
Outcome: Subprime approval        
  Black 2.559 (0.022)*** 2.415 (0.023)*** 0.731 (0.038)*** 2.576 (0.039)***
  Latino 2.601 (0.022)*** 2.450 (0.022)*** 0.966 (0.051) 2.457 (0.036)***
  Asian 0.624 (0.011)*** 0.723 (0.012)*** 0.067 (0.008)*** 0.615 (0.018)***
  Other 1.462 (0.057)*** 1.609 (0.063)*** 1.648 (0.065)*** 1.614 (0.063)***
  No race identified 1.191 (0.021)*** 1.165 (0.021)*** 1.165 (0.021)*** 1.165 (0.021)***
  Multiracial 0.875 (0.042)** 0.855 (0.042)** 0.858 (0.042)** 0.853 (0.041)**
  Female 0.944 (0.005)*** 0.920 (0.005)*** 0.917 (0.005)*** 0.920 (0.005)***
  Log (income) 0.625 (0.003)*** 1.024 (0.006)*** 0.941 (0.006)*** 1.024 (0.006)***
  Has co-applicant   0.651 (0.004)*** 0.658 (0.004)*** 0.651 (0.004)***
  Log (loan amount)   0.780 (0.005)*** 0.780 (0.005)*** 0.779 (0.005)***
  Tract percentage of people    of color   0.998 (0.000)*** 0.999 (0.000)*** 0.998 (0.000)***
  Log (tract income)   0.412 (0.006)*** 0.413 (0.006)*** 0.413 (0.006)***
  In MSA/MD   0.681 (0.008)*** 0.690 (0.008)*** 0.681 (0.008)***
  Mid-Atlantic   1.027 (0.022) 1.013 (0.021) 1.024 (0.022)
  East North Central   1.357 (0.026)*** 1.343 (0.026)*** 1.357 (0.026)***
  West North Central   1.481 (0.033)*** 1.465 (0.032)*** 1.480 (0.033)***
  South Atlantic   1.176 (0.023)*** 1.167 (0.022)*** 1.173 (0.023)***
  East South Central   1.546 (0.035)*** 1.540 (0.035)*** 1.543 (0.035)***
  West South Central   1.787 (0.037)*** 1.793 (0.037)*** 1.779 (0.037)***
  Mountain   1.113 (0.026)*** 1.113 (0.026)*** 1.112 (0.026)***
  Pacific   1.286 (0.027)*** 1.253 (0.026)*** 1.277 (0.027)***
  Black × log income     1.337 (0.017)***  
  Latino × log income     1.243 (0.015)***  
  Asian × log income     1.699 (0.044)***  
  Black × neighborhood's    percentage of nonwhites       0.999 (0.000)***
  Latino × neighborhood's    percentage of nonwhites       1.000 (0.000)
  Asian × neighborhood's    percentage of nonwhites       1.004 (0.001)
  Constant 0.607 (0.014)*** 6,201.882 (909.609)*** 8,493.775 (1,238.971)*** 6,120.317 (899.170)***
Observations 3,341,693 3,340,428 3,340,428 3,340,428
Pseudo R 2 .044 .066 .067 .066

 

Description: 

Table 2 presents results from the four multinomial logits where prime approval is the base outcome. The top panel compares the outcome of loan denial with prime approval, and the bottom panel compares subprime approval with prime approval. Because the findings are odds ratios, each coefficient should be interpreted as the effect a one-unit change in the covariate has on the odds of the corresponding outcome being the result as opposed to the outcome of prime mortgage approval.88. If a coefficient is greater than 1, an increase in the variable increases the odds. If a coefficient is less than 1, an increase in the variable decreases the odds.View all notes 

For example, comparing black applicants to identical white applicants with respect to the measures controlled for in model 1, outcomes of their applications differed substantially: Blacks were 3.6 times more likely than whites to be denied a loan, and if approved, they were 2.6 times more likely to be offered a subprime loan rather than a prime loan. The first part of this section will discuss the basic, noninteracted models (models 1 and 2). The second part will consider the borrower race interactions with income and neighborhood racial composition (models 3 and 4). This section focuses on the primary variables of interest: borrower race, borrower income, neighborhood racial makeup, and interaction terms of race with income and race with neighborhood racial makeup.

 

 

Thanks for posting this information, however I don’t see anything in this data specifically regarding the quote in the OP.

Link to comment
Share on other sites

1 hour ago, Incredulity said:

Thanks for posting this information, however I don’t see anything in this data specifically regarding the quote in the OP.

The author just plugged in the numbers (income=200k, race=black) into one of the models on the bottom of the table.

Link to comment
Share on other sites

1 hour ago, TwiceHorn said:

The Center for Responsible Lending found that during the housing boom, 6.2 percent of whites with a credit score of 660 and higher received high-interest mortgages but 21.4 percent of blacks with a score of 660 or higher received these same loans.

These are exactly the kind of statistics that are super easy to manipulate. A credit score of 660 won’t get you a prime mortgage. The cutoff for a prime loan is substantially higher. Although the statement is likely true, and certainly appears to damning, it’s misleading. 

2 hours ago, TwiceHorn said:

It turned out that several of the major banks had been purposely giving people of color subprime mortgages, including borrowers who would have qualified for a prime loan.

As I mentioned in a prior post, people (regardless of race) should shop around  for the best mortgage terms. The people making loans are on the other side of the table. It’s unfortunate that many people, especially people of color it seems, are ignorant to this fact. 

2 hours ago, TwiceHorn said:

One former employee testified that in 2001, Wells Fargo created a unit that would be responsible for pushing expensive refinance loans on black customers, especially those living in Baltimore, southeast Washington, D.C., and Prince George’s County—all locations with large black populations. 

A sales organization with geographic territories?!?!?!?! Shocking. Wait. This describes countless other sales organizations. 

2 hours ago, TwiceHorn said:

Specifically targeted for subprime loans among the minority demographic were black women. Women of color are the most likely to receive subprime loans

Have you read the Wives and the Stupid Things They Do Thread????

Lastly,  I heard Madoff’s sales organation targets mostly old, rich white people. Fucking racists. 

Bernard

  • Fuck You 1
Link to comment
Share on other sites

I have little doubt that housing policy (FHA) and housing practices (mortgage lending, restrictive covenants) have been discriminatory.  There's too much evidence to ignore it.

It does blow my mind that a commercial outfit in the last decade, e.g. Wells Fargo, would be so blatantly discriminatory.  It just baffles me that some right-minded employees didn't blow the whistle on it.

What is kind of intriguing, though, is the current notion that housing is a poor investment, documented elsewhere on the board:  i.e. that most people would have been better off investing the delta between renting and home ownership and that the US policy of pushing home ownership has been a mistake.  On the one hand, the plight of black folks with odds intentionally stacked against them in regards to home ownership would seem to belie this notion.  On the other, it seems the financial fragility of black households in the face of the mortgage crisis (all the stats showing how much worse off black families were after) maybe supports the notion.

Link to comment
Share on other sites

17 hours ago, Catdaddyhorn said:

Does this count as institutional racism?

 

This is absolutely fucking criminal.  

I'm always (more than) a bit skeptical at claims like this when no data is given to support it. (not saying that I'm not aware of this kind of red-lining crap going on).

Show me the data and let me come to my own conclusions. What exactly does "on average" mean here?

But yeah, this is institutional racism. It's bad. My point is show me the data so I can determine how bad it is. 

Link to comment
Share on other sites

I have no doubt (see: Wells Fargo) that discriminatory and predatory lending practices exist.

I also have no doubt that some minority groups have been victimized by said lending practices.

That being said, OP provides a tweet and expects us to all nod in agreement.

There are so many factors at play here. Just because something or some things happened in the ether does not exactly equal a certain conclusion and, more specifically, is not the cause of that certain conclusion.




Sent from my iPhone using Tapatalk Pro

Link to comment
Share on other sites

1 hour ago, Asithappens said:

I'm always (more than) a bit skeptical at claims like this when no data is given to support it. (not saying that I'm not aware of this kind of red-lining crap going on).

Show me the data and let me come to my own conclusions. What exactly does "on average" mean here?

But yeah, this is institutional racism. It's bad. My point is show me the data so I can determine how bad it is. 

Warning ... about to make a blanket statement. Agree 100%. From my very limited POV, much of broad social media outrage is not supported by data or analytics. On occasion, when one does find an underlying study, often when one digs into the details, either it does not support the conclusion or the way it was done was designed to reveal a pre-conceived result. This feels like a Smollett - make a wild ass claim and then demand some payoff to go away or get a concession.

Link to comment
Share on other sites

What I have seen personally as a lawyer, that relates to this, is financial ignorance.  Using the term ignorance in the most benign way possible:  a simple lack of knowledge.  Because blacks (and others) have been deliberately (in many cases) placed behind the 8 ball, their generational families don't acquire familial knowledge of mortgages, decent mortgage rates and terms, car loan rates and terms, making simple Boglehead investments, and other personal financial issues.  This applies with equal force to all poors, but unlike being black, poors may have relatives or close acquaintances or neighbors that aren't poor or are at least experienced and could give them some guidance on these issues.

An issue I have seen several times and know it to be a broader issue is that black/poor families that do manage to acquire real property don't know to probate it on death and so the chain of title to the family farm or homestead, or even Uncle Larry's place, is fucked-up beyond all recognition and no one in the family is able to receive anything close to market value for it.  You can find this about to go down on the CYHMWT board all the time, so it's an ignorance issue not a race or racial ignorance thing.  But, by "institutional" and "structural" racism, we more or less insure(d) that black families would be less likely to be property owners and thus even less likely that they would be "smart" property owners.

There are dozens of little things like this that aggregate to keep blacks and poors "down," even when they are doing most everything else right.  It's a helluva problem.

Edited by TwiceHorn
  • Like 1
Link to comment
Share on other sites

1 minute ago, BurntEyes said:

I'll just add, I think the political atmosphere and climate, exacerbated by certain segments of the population have actually increased racial tensions in this country. They've done so by latching on to many various activities and slapping the racist sticker on it.

I think it's an intentional, Machiavellian play to drive the voters even further apart rather than addressing the issues as a national problem.

I submit predatory lending practices as one such example and police abuses of power under the auspice of the WOD as another.

These are US issues, not race issues.

I have said for quite a few years now that the racial dialog in this country sucks.  But, that may be the nature of the racial dialog and it may be necessary in order to get past it.

Link to comment
Share on other sites

57 minutes ago, BurntEyes said:

Are there predatory lending practices by mortage company?

Absolutely.

Are uniformed people of all races predated upon?

Absolutely.

Are certain subsets of the population likely to be predated upon at ah higher rate due to both habit of lenders and lack of knowledge on the buyers part?

Absolutely.

Is it institutional racism?

Not nessecarily.

It's a predator being a predator.

A predator chases the easy pray. 

 

 

If the answer to your first three questions is yes, then you have, by definition, institutional racism.

Link to comment
Share on other sites

1 minute ago, po elvis said:

Exactly what are the things that it takes to be the greatest country on earth? and which country would that be?

Well, for starters it can't be a white nation, we know that much.

From there everything must be given away free to the citizens, and they all get too ride multi colored flying uni-corns.

Link to comment
Share on other sites

2 hours ago, Asithappens said:

I'm always (more than) a bit skeptical at claims like this when no data is given to support it. (not saying that I'm not aware of this kind of red-lining crap going on).

Show me the data and let me come to my own conclusions. What exactly does "on average" mean here?

But yeah, this is institutional racism. It's bad. My point is show me the data so I can determine how bad it is. 

"On average" means that all the other variables have been accounted for. Variables like, oh I don't know, credit scores. It means that a black family and a white family applying for a loan, all other variables being the same, the black family was more likely to get a sub-prime loan, despite having a higher income.

 

And lulz at "show me the data". Are you really going to look at hundreds, probably more like thousands of data points, and do the regression analysis yourself?

 

 

Link to comment
Share on other sites

2 minutes ago, High Plains Drifter said:

"On average" means that all the other variables have been accounted for. Variables like, oh I don't know, credit scores. It means that a black family and a white family applying for a loan, all other variables being the same, the black family was more likely to get a sub-prime loan, despite having a higher income.

And lulz at "show me the data". Are you really going to look at hundreds, probably more like thousands of data points, and do the regression analysis yourself.

Give this thread enough time and people will start to rationalize red lining as sound business practice.

Link to comment
Share on other sites

15 hours ago, Bernard said:

It’s not racism.

I got a loan to buy a house in 2004 from Countrywide Home Loans, then largest home mortgage lender in the county. I did the whole thing by phone/fax/mail. I doubt they knew if I was black or white until long after the loan terms were settled. 

 

Bernard

 

They probably didn't know you were black or white, but based on your address and other factors they probably had a pretty good idea you were white (or black).

Link to comment
Share on other sites

1 hour ago, TwiceHorn said:

What I have seen personally as a lawyer, that relates to this, is financial ignorance.  Using the term ignorance in the most benign way possible:  a simple lack of knowledge.  Because blacks (and others) have been deliberately (in many cases) placed behind the 8 ball, their generational families don't acquire familial knowledge of mortgages, decent mortgage rates and terms, car loan rates and terms, making simple Boglehead investments, and other personal financial issues.  This applies with equal force to all poors, but unlike being black, poors may have relatives or close acquaintances or neighbors that aren't poor or are at least experienced and could give them some guidance on these issues.

An issue I have seen several times and know it to be a broader issue is that black/poor families that do manage to acquire real property don't know to probate it on death and so the chain of title to the family farm or homestead, or even Uncle Larry's place, is fucked-up beyond all recognition and no one in the family is able to receive anything close to market value for it.  You can find this about to go down on the CYHMWT board all the time, so it's an ignorance issue not a race or racial ignorance thing.  But, by "institutional" and "structural" racism, we more or less insure(d) that black families would be less likely to be property owners and thus even less likely that they would be "smart" property owners.

There are dozens of little things like this that aggregate to keep blacks and poors "down," even when they are doing most everything else right.  It's a helluva problem.

Based on what I hear second hand from my housing attorney wife, this is exactly correct. And often the lender, whether it be Wells Fargo or Uncle Larry, deliberately take advantage of the ignorance of the borrower.

If the ignorant borrowers are predominantly black (or Hispanic or whatever), I'd say that amounts to "institutional racism".

Link to comment
Share on other sites

1 minute ago, High Plains Drifter said:

Based on what I hear second hand from my housing attorney wife, this is exactly correct. And often the lender, whether it be Wells Fargo or Uncle Larry, deliberately take advantage of the ignorance of the borrower.

If the ignorant borrowers are predominantly black (or Hispanic or whatever), I'd say that amounts to "institutional racism".

But is that intentional racism or just luck of the draw for the institution ? They'd take advantage of the dumb white people as well I'd imagine.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.



×
×
  • Create New...