Hispanic Homeownership Seminar
-
Upload
gary-oconnor -
Category
Documents
-
view
28 -
download
1
description
Transcript of Hispanic Homeownership Seminar
Hispanic Homeownership Seminar
U.S. Department of Housing and Urban DevelopmentThursday, June 1, 2006
Hispanic Homeownership Seminar - June 1, 2006 2
Housing Tenure Expenditure and Satisfaction across Hispanic, African-
American, and White Households: Evidence from the
American Housing Survey
Thomas P. Boehm
and
Alan M. Schlottmann
Hispanic Homeownership Seminar - June 1, 2006 3
Two Primary Components to the Analysis of the Differencesin Housing Circumstances across Racial/Ethnic Groups
• Housing Tenure and Housing Expenditure
• Quality of Structural and Neighborhood Environment
Hispanic Homeownership Seminar - June 1, 2006 4
The Data
Pooled recent AHS MSA samples:
• 41 MSAs – sampling years 1998, 2002, and 2004
• Approximately 5,000 households in each MSA
Why the MSA Samples?
• 17,968 Hispanic Households in the combined sample
• Identify the housing markets in which housing decisions are being made
Hispanic Homeownership Seminar - June 1, 2006 5
I. Housing Tenure and Expenditure
Two Sub-samples Considered in the Analysis
• All Households
• Recent Movers
Stratification of the Samples
• High versus Low Income
• Non-White versus White Hispanics
• Recent Immigrant Status
Hispanic Homeownership Seminar - June 1, 2006 6
The Results
Analysis of Means
• Ownership Is Substantially More Likely for Whites
Low-income Hispanics and African-Americans ownership rates for the full sample are 38.0 % and 36.5 % respectively versus 60.8 % for whites.
Hispanic Homeownership Seminar - June 1, 2006 7
• House Value and Rental Cost Substantially Higher for White Households
Low Income Households – Full Sample
Mean House Value:
Whites - $147,289 African Americans - $95,055Hispanics - $128,681
Mean Rent Payment:
Whites - $606African Americans - $496Hispanics - $592
Hispanic Homeownership Seminar - June 1, 2006 8
Differences for Hispanic Households
1. Hispanic homeowners experience higher total amounts of debt:
For example, the full sample debt levels for low income houses:
White households – $35,509Black households – $35,164Hispanic households – $45,971
Hispanic Homeownership Seminar - June 1, 2006 9
Differences for Hispanic Households
2. Hispanic Households are more crowded than other groups
For example, for the full sample for lower income families:
Owners – Hispanics average 681 square feet per person, while African-Americans and whites average 1,062 and 1,130 square feet per person respectably.
Renters – Hispanics average 391 square feet per person, African-Americans 574, whites 640.
Hispanic Homeownership Seminar - June 1, 2006 10
Regression Analysis
• Likelihood of Homeownership – Is the likelihood of homeownership lower for different racial/ethnic groups when controlling for other factors that might influence the likelihood that they would be able to purchase a home?
• Control factors – The housing market (41 MSAs), the marital status and sex of household head, age of the household head, the family size, the income of the family and a categorical variable identifying families that have substantial savings ($25,000 or more).
Hispanic Homeownership Seminar - June 1, 2006 11
Both minority groups have a substantially lowerlikelihood of ownership
• For the full sample, for low income families
African-Americans 40.5 % less likely to ownWhite-Hispanics 28.5 % less likely to ownNon-White Hispanics 39.2 % less likely to own
• For the recent movers, for low income families:
African-Americans 52.3 % less likely to ownWhite Hispanics 38.2 % less likely to ownNon-White Hispanics 23.0 % less likely to own
Hispanic Homeownership Seminar - June 1, 2006 12
Differential Levels of House Value for Homeowners
Controlling for other factors as before(i.e., income, education, the housing market, etc.)
• For the full sample for low-income owners:
African-Americans $20,113 lower house value White-Hispanics $19,768 lower house valueNon-White Hispanics $28,845 lower house value
• For the recent mover sample for low-income owners:
African-Americans $35,077 lower house value White-Hispanics $23,715 lower house valueNon-White Hispanics $45,019 lower house value
Hispanic Homeownership Seminar - June 1, 2006 13
Differential Annual Rent Payment for Renters
• For the Full Sample of low-income renters:
African-American Families $712 lower rent levelsWhite-Hispanic Families $764 lower rent levelsNon-white Hispanic Families $871 lower rent levels
• For the Recent-mover sample of low-income renters:
African-American Families $742 lower rent levelsWhite-Hispanic Families $844 lower rent levelsNon-white Hispanic Families $946 lower rent levels
Hispanic Homeownership Seminar - June 1, 2006 14
The Impact of “Time in the U.S.” on Homeownership, and House Value and Rent Levels
• Available only for 2002 and 2004
• Categorical representation of “time in the U.S.”
– 5 years or less
– 5 through 12 years
– 13 through 22 years
– 23 years or more
Hispanic Homeownership Seminar - June 1, 2006 15
Results for Recent Immigrants
• Households in the U.S. for less than 5 years and 5-12 years are significantly less likely to buy a home than households whose head was born in this country holding other factors constant.
• In contrast, for the recent mover sample these recent immigrant groups do not have a substantially lower likelihood of being owners
• For both groups, renters have a significantly lower, rental expenditure but there is no impact of recent immigrant status on the value of house owned by either low or high income households.
Hispanic Homeownership Seminar - June 1, 2006 16
II. Satisfaction with Structural and Neighborhood Characteristics
AHS – 10 point scale ranking structural and neighborhooddimension of housing units.
Mean Ranking
• The primary determinant of these rankings is tenure – owners rank both neighborhood and structural characteristics higher than renters
Hispanic Homeownership Seminar - June 1, 2006 17
Low-Income – Full Sample
Owners
Structural Rank Neighborhood RankWhite 8.44 8.14Black 8.31 7.71Hispanic 8.36 8.02
Renters
Structural Rank Neighborhood RankWhite 7.46 7.55Black 7.41 7.33Hispanic 7.49 7.54
Hispanic Homeownership Seminar - June 1, 2006 18
Analysis of Individual Structural and Neighborhood Characteristics
• Ordinal Probit Model
• Equations Stratified by Tenure, Race/Ethnic Group, and Income
Hispanic Homeownership Seminar - June 1, 2006 19
Results for Structural Quality
All households react in a very similar manner tostructural characteristics.
• Factors lowering Structural Quality: External leaks, internal leaks, vermin problems, major structural problems, and interior deterioration, etc.
• Factors increasing Structural Quality: Being on a public sewer system, plumbing and toilets being in satisfactory working order, having a central heating system, etc.
Hispanic Homeownership Seminar - June 1, 2006 20
Results for Structural Quality
Not all groups have identical characteristicscircumstances
• For the full sample, minorities do have a higher proportion with one or more significant structural problem.
Low-Income Owned Units – Full Sample
Hispanics 43.2% with major structural problems
African-Americans 54.4% with major structural problems
Whites 23.0% with major structural problems
Hispanic Homeownership Seminar - June 1, 2006 21
• Water quality issues are worse for Hispanic households than African American or white households
Low-Income Homeowners
Hispanic 24.3% water quality a problem
African-American 13.4% water quality a problem
White 8.5% water quality a problem
Hispanic Homeownership Seminar - June 1, 2006 22
• Heating system quality is generally worse for Hispanic households than African-Americans or White households
Low-Income Homeowners
Hispanic 10.7% water quality a problem
African-American 6.3% water quality a problem
White 2.8% water quality a problem
Hispanic Homeownership Seminar - June 1, 2006 23
• Low-income Hispanic households face more crowding than do their African-American or white counterparts.
Low-Income Homeowners
Hispanic 2.5 rooms per person
African-American 3.5 rooms per person
White 3.8 rooms per person
Hispanic Homeownership Seminar - June 1, 2006 24
• Deterioration of Interior facilities (i.e., cracks, holes in the walls or ceiling, holes in the floor, broken plaster or peeling paint)
Low-Income Homeowners
Hispanic 9.2%
African-American 11.8%
White 6.0%
Hispanic Homeownership Seminar - June 1, 2006 25
Results for Neighborhood Quality
All households react in a similar manner to factorseffecting neighborhood quality
• Factors Making a Neighborhood Undesirable: crime problems, litter problems, noise problems, junk and abandoned buildings, etc.
• Factors Making a Neighborhood Desirable: Green space, newer buildings, etc.
Hispanic Homeownership Seminar - June 1, 2006 26
Results for Neighborhood Quality
Not all groups have identical characteristics circumstances
• Crime problems and inadequate police protection are more likely to be perceived by African-Americans and White Households.
Low-Income Renters
Hispanic 8.9% Police Protection Inadequate
African-American 15.9% Police Protection Inadequate
White 6.1% Police Protection Inadequate
Hispanic Homeownership Seminar - June 1, 2006 27
• Green space is less likely to be near minority owned homes.
Low-Income Owners
Hispanic 23.79% Green Space Nearby
African-American 27.18% Green Space Nearby
White 35.67% Green Space Nearby
Hispanic Homeownership Seminar - June 1, 2006 28
• African-Americans owners and renters are more likely to have abandoned buildings nearby.
Low-Income Renters
Hispanic 8.9% Abandoned Buildings a problem
African-American 15.9% Abandoned Buildings a problem
White 6.1% Abandoned Buildings a problem
Hispanic Homeownership Seminar - June 1, 2006 29
• Minority renters are more likely to be located in areas in which road repairs are a concern.
Low-Income Renters
Hispanic 43.4% Road repairs a problem
African-American 50.3% Road repairs a problem
White 38.0% Road repairs a problem
Hispanic Homeownership Seminar - June 1, 2006 30
Summary
1. First time the housing circumstances of a large national sample of Hispanics compared to African-Americans and Whites.
2. A primary determinant of structural and neighborhood satisfaction is homeownership. Clearly, Hispanics and African-Americans are still at a disadvantage.
3. The lower value of rented and owned units for African-American households in the sample suggests that they might be considered generally worse off than Hispanics with regard to housing
4. Regardless of Racial/Ethnic background the same specific neighborhood and structural factors influence housing satisfaction, but characteristics of the housing in which they live can be quite different even within a particular income group and tenure type.
Hispanic Homeownership Seminar - June 1, 2006 31
Summary
5. Recent Immigrant Status can have a significant on housing choice influence.
6. Some unique aspects of the Hispanic housing situation were observed
a. White and Non-White Hispanics often appeared to have significantly worse outcomes, even when other factors were controlled for
b. Hispanic households appear generally more crowded
c. Hispanic households appear to have relatively high home mortgage debt levels
d. Hispanic households are more likely to experience problems with water quality and live in units with lower quality heating sources
Hispanic Homeownership Seminar - June 1, 2006 32
Homeownership Rate Differences Between Hispanics andNon-Hispanic Whites:
Regional Variation at the County Level
George S. Masnick
Hispanic Homeownership Seminar - June 1, 2006 33
Research Goals
• Better understand the magnitude of homeownership rate gaps between Hispanics and non-Hispanic whites.
• Better understand the relationships between homeownership rate gaps and selected explanatory variables.
• Identify counties that are regional outliers, where homeownership rate gaps are large by regional standards but where explanatory differences are weak.
• Such counties could be targets of future efforts to improve homeownership opportunities for young Hispanic families.
Hispanic Homeownership Seminar - June 1, 2006 34
Background Facts
• Hispanics are geographically concentrated.
• Hispanic household heads are younger.
• Non-Hispanic white/ Hispanic home ownership rate gaps are smallest for younger householders.
• West and South have smaller home ownership rate gaps than Midwest and Northeast
Hispanic Homeownership Seminar - June 1, 2006 35
Methodology
• Select the 25 counties in each census region that have the highest numerical Hispanic representation.
• Focus on young adults 25-34.
• Maintain a regional focus.
• Data source is 2000 census summary files
• Rely on graphic representation to understand Hispanic/ non-Hispanic white differences (Chart Type #1).
Hispanic Homeownership Seminar - June 1, 2006 36
Chart Type #1 - West
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Hispanic vs White Home Ownership RatesWest: Owners Age 25-34
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Hispanic HO Rate
NH
Wh
ite H
O R
ate
Dona Ana, NM
Maricopa, AZ
Orange, CABemalillo, NM
Fresno, CA
Ventura, CA
San Francisco, CA
Monterey, CA
Riverside, CA
San Mateo, CA
Figure 1b
No
n-H
isp
an
ic W
hit
e H
o R
ate
Hispanic Homeownership Seminar - June 1, 2006 37
Chart Type #1 - South
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Hispanic vs White Home Ownership RatesSouth: Owners Age 25-34
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Hispanic HO Rate
NH
Wh
ite H
O R
ate
Fort Bend, TXGwinnett, GA
DeKalb, GA
Brazoria, TXPalm Beach, FL
Hidalgo, TX
Webb, TX
El Paso TX
Denton TX
Prince Georges, MD
Figure 2bNo
n-H
isp
an
ic W
hit
e H
o R
ate
Hispanic Homeownership Seminar - June 1, 2006 38
Chart Type #1 - Midwest
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Hispanic vs White Home Ownership RatesMidwest: Owners Age 25-34
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Hispanic HO Rate
NH
Wh
ite H
O R
ate
Finney, KS
Will, IL
McHenry, IL
Johnson, KS
Hennepin, MN
Marion, IN
Kent, MiLorain, OH
Oakland, MIWayne, MI
Figure 3bNo
n-H
isp
an
ic W
hit
e H
o R
ate
Hispanic Homeownership Seminar - June 1, 2006 39
Chart Type #1 - Northeast
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Hispanic vs White Home Ownership RatesNortheast: Owers Age 25-34
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Hispanic HO Rate
NH
Wh
ite H
O R
ate Nassau, NY
Fairfield, CT
Philadelphia, PA
Passaic, NJHartford, CT
Hampden, MAWorcester, MA
Figure 4bNo
n-H
isp
an
ic W
hit
e H
o R
ate
Hispanic Homeownership Seminar - June 1, 2006 40
Explanatory Variables
• Median Income and % Age 25-34 Earning $40K+
• Median Housing Value
• Hispanic Housing Costs as a % of Household Income
• Share Owner Units Built Since 1980
• Share Owner Units Single Family Detached
• Share of Hispanic Foreign Born Not a Citizen
• Share Age 15-64 Married with Kids <18
Hispanic Homeownership Seminar - June 1, 2006 41
Explanatory Variables
• Median Income and % Age 25-34 Earning $40K+
• Median Housing Value
• Housing Costs as a % of Household Income
• Share Owner Units Built Since 1980
• Share Owner Units Single Family Detached
• Share of Hispanic Foreign Born Not a Citizen
• Share Age 15-64 Married with Kids <18
Hispanic Homeownership Seminar - June 1, 2006 42
Median Housing Value - West
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Housing Value Owner Occupied Units
West
$0
$100,000
$200,000
$300,000
$400,000
$500,000
$600,000
$0 $200,000 $400,000 $600,000
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers
San Mateo, CA
Santa Clara, CA
San Francisco, CA
Santa Barbara, CA
Monterey, CA
Alameda, CA
Figure 8a
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 43
Median Housing Value - South
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Housing Value Owner Occupied Units
South
$0
$50,000
$100,000
$150,000
$200,000
$250,000
$300,000
$0 $100,000 $200,000 $300,000
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers Fairfax, VA
Montgomery, MD
DeKalb, GA
Figure 8b
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 44
Median Housing Value - Midwest
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Housing Value Owner Occupied Units
Midwest
$0
$50,000
$100,000
$150,000
$200,000
$250,000
$300,000
$0 $100,000 $200,000 $300,000
Hispanic Owners
Non-H
ispanic
Whit
e O
wners
Lake, IL
Wyandotte, KS
Kane IL
DuPage, IL
Oakland, MI
Figure 8c
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 45
Median Housing Value - Northeast
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Housing Value Owner Occupied Units
Northeast
$0
$50,000
$100,000
$150,000
$200,000
$250,000
$300,000
$0 $100,000 $200,000 $300,000
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers
Excluding New York County = $245K Hispanic, $1,000K White
Philadelphia, PA
Camden, NJ Hudson, NJ
Essex, MA
Essex, NJ
Kings, NYQueens, NY
Figure 8d
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 46
Explanatory Variables
• Median Income and % Age 25-34 Earning $40K+
• Median Housing Value
• Housing Costs as a % of Household Income
• Share Owner Units Built Since 1980
• Share Owner Units Single Family Detached
• Share of Hispanic Foreign Born Not a Citizen
• Share Age 15-64 Married with Kids <18
Hispanic Homeownership Seminar - June 1, 2006 47
Median Cost Burden - West
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Owner Costs as a % of Income Owners with a Mortgage
West
15
20
25
30
35
15 20 25 30 35
Hispanic Owners
No
n-H
isp
an
ic W
hit
e O
wn
ers
Dona Ana, NM
Kern, CA
San Francisco, CA
Los Angeles, CA
Figure 10a
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 48
Median Cost Burden - South
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Owner Costs as a % of Income Owners with a Mortgage
South
15
20
25
30
35
15 20 25 30 35
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers
Brazoria, TX
Miami-Dade, FL
Orange, FL
Figure 10b
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 49
Median Cost Burden - Midwest
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Owner Costs as a % of Income Owners with a Mortgage
Midwest
15
20
25
30
35
15 20 25 30 35
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers
McHenry, IL
Wyandotte, KS
Figure 10c
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 50
Median Cost Burden - Northeast
• Diagonal represents equality
• Short dark line represents trend
• Separation between lines represents departure from equality
• Points above trend line represent outliers
Median Owner Costs as a % of Income Owners with a Mortgage
Northeast
15
20
25
30
35
15 20 25 30 35
Hispanic Owners
Non
-His
pan
ic W
hit
e O
wn
ers
Philadelphia, PA
New York, NY
Queens, NY
Figure 10d
No
n-H
isp
an
ic W
hit
e O
wn
ers
Hispanic Homeownership Seminar - June 1, 2006 51
Methodology
• Rely on graphic representation to study the importance of explanatory variables and to identify outliers (Chart Type # 2).
Hispanic Homeownership Seminar - June 1, 2006 52
Chart Type #2 (Median Value) - West
• Explanatory variables generally show weak relationships with home ownership gaps
• Outliers of interest are those above the line of best fit and to the left
Correlation Between Median Value and 25-34 HO Gap - West
R2 =
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
$0 $100,000 $200,000 $300,000 $400,000 $500,000
Median Housing Value
Ho
me
Ow
ne
rsh
ip R
ate
Gap
San Mateo, CAOrange, CAMaricopa, AZ
Ventura, CA
Fresno, CA
Clark, NV
Figure 9a
Ho
me
Ow
ne
rsh
ip R
ate
Ga
p
Hispanic Homeownership Seminar - June 1, 2006 53
Chart Type #2 (Median Value) - South
• Explanatory variables generally show weak relationships with home ownership gaps
• Outliers of interest are those above the line of best fit and to the left
Correlation Between Median Value and 25-34 HO Gap - South
R2 =
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
$0 $100,000 $200,000 $300,000 $400,000 $500,000
Median Housing Value
Ho
me
Ow
ne
rsh
ip R
ate
Gap
Gwinnett, GA
DeKalb, GA
Prince Georges, MD
Oklahoma, OK Denton TX Fairfax, VA
Montgomery, MD
Figure 9b
Ho
me
Ow
ne
rsh
ip R
ate
Ga
p
Hispanic Homeownership Seminar - June 1, 2006 54
Chart Type #2 (Median Value)- Midwest
• Explanatory variables generally show weak relationships with home ownership gaps
• Outliers of interest are those above the line of best fit and to the left
Correlation Between Median Value and 25-34 HO Gap - Midwest
R2 =
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
$0 $100,000 $200,000 $300,000 $400,000 $500,000
Median Housing Value
Ho
me
Ow
ne
rsh
ip R
ate
Gap
McHenry, ILMarion, IN
Johnson, KSHennepin, MN
Kent, Mi
Lorain, OH
Oakland, MIWayne, MI
Figure 9c
Ho
me
Ow
ne
rsh
ip R
ate
Ga
p
Hispanic Homeownership Seminar - June 1, 2006 55
Chart Type #2 (Median Value) - Northeast
• Explanatory variables generally show weak relationships with home ownership gaps
• Outliers of interest are those above the line of best fit and to the left
Correlation Between Median Value and 25-34 HO Gap - Northeast
R2 =
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
$0 $100,000 $200,000 $300,000 $400,000 $500,000
Median Housing Value
Ho
me
Ow
ne
rsh
ip R
ate
Gap Nassau, NY
Fairfield, CTHampden, MA
Passaic, NJWorcester, MA
Hartford, CT
New Haven, CT
Not Including New York County Median Value = $1,000,001 Gap = 10.2%
Figure 9d
Ho
me
Ow
ne
rsh
ip R
ate
Ga
p
Hispanic Homeownership Seminar - June 1, 2006 56
Outliers Short List (Criteria)
• Must be Ranked in Top 10 (among 25 Counties) on Largest Home Ownership Rate Gaps for 25-34 Year olds in Each Region
• Pick Five Counties Down the Top 10 Gap List that Score 15 or Lower on Average Ranking of All Seven Explanatory Variables Combined
Hispanic Homeownership Seminar - June 1, 2006 57
Outliers Short List (Counties)
• West -– Maricopa, AZ– Fresno, CA– Clark, NV– San Joaquin, CA– Contra Costa, CA
• South -– DeKalb, GA– Prince George’s, MD– Denton, TX– Palm Beach, FL– Oklahoma, OK
• Midwest -– Johnson, KS– Marion, IN– McHenry, IL– Lorain, OH– Kent, MI
• Northeast -– Worcester, MA– Hartford, CT– Hampden, MA– Nassau, NY– New Haven, CT
Hispanic Homeownership Seminar - June 1, 2006 58
Summary
Summary and Conclusions…
Hispanic Homeownership Seminar - June 1, 2006 59
Mortgage Pricing Differentials Across Hispanic, African-American, and White
Households: Evidence from the American Housing Survey
Thomas P. Boehm
and
Alan M. Schlottmann
Hispanic Homeownership Seminar - June 1, 2006 60
Literature on Racial/Ethnic Differences in Mortgage Lending
• Mortgage Origination and Redlining
• Mortgage Pricing
• Mortgage Pricing – Using the American Housing Survey
– Susin – 2003
– Boehm/Schlottmann/Thistle – 2004
Hispanic Homeownership Seminar - June 1, 2006 61
Comparison to the Literature
• Larger Sample of Hispanic Households
• Markets in which loans were originated can be identified as well as the year in which they were originated
• More stratification can be done by:
– Minority subgroup
– Loan type
» Conventional versus VA/FHA
» Refinancing versus Home Purchase
» First Mortgage versus junior mortgage and home equity loans
Hispanic Homeownership Seminar - June 1, 2006 62
The Data
Pooled recent AHS MSA samples:
• 41 MSAs – sampling years 1998, 2002, and 2004
• Mortgages of Different Type in the Sample
First Mortgages - Full Sample
High Income Low Income
HispanicAfrican
American White HispanicAfrican
American White Totals
Conv / Home Purchase 960 967 1,717 1,792 1,261 5,221 11,918
Gov. / Home Purchase 495 797 631 861 1,016 1,588 5,388
Conv / Refinance 283 280 725 794 329 2,955 5,366
Gov. / Refinance 83 74 118 203 165 446 1,089
23,761
Hispanic Homeownership Seminar - June 1, 2006 63
The Data
First Mortgages – Recent Mover Sample
High Income Low Income
HispanicAfrican
American White HispanicAfrican
American White Totals
Conv / Home Purchase 241 198 367 528 329 1,334 2,997
Gov. / Home Purchase 153 172 155 249 231 381 1,341
4,338
Junior Mortgages – Full Sample
High Income Low Income
HispanicAfrican
American White HispanicAfrican
American White Totals
Junior Mortgage 102 174 252 414 259 1,033 2,234
Home Equity Loan 74 80 393 283 147 1,493 2,470
4,704
Hispanic Homeownership Seminar - June 1, 2006 64
Constraints on the Sample of 1st Mortgages
• Mortgages made by financial institution
• Mortgage made on residence only
• Unit not condominium or manufactured
• Loan not assumption or wraparound
• For 1st Mortgages – loan terms 10, 15, 20, 25, or 30 years
Hispanic Homeownership Seminar - June 1, 2006 65
Analysis of First Mortgages
Average Mortgage Interest Rates
• Interest rates are substantial higher for low-income families
• Hispanics higher than African-Americans in all but one market segment (conventional refinance, full sample)
Recent Movers – Conventional – Purchase
Low-Income Hispanic 7.20%
Low-Income African-American 6.81%
Low-Income White 6.63 %
Hispanic Homeownership Seminar - June 1, 2006 66
Regression Analysis
Model Specification
• Three categories of control variables:
– Characteristics of the borrower
– Characteristics of the property
– Characteristics of the loan itself
• Cannot determine if discrimination exists because data does not include:
– Individual’s net-wealth or credit history
– Institution’s characteristics or underwriting criteria
Hispanic Homeownership Seminar - June 1, 2006 67
Home Purchase Market – Effects of Race/Ethnicity
• In the conventional market for the full sample Non-white and white Hispanics pay significantly higher rate than whites (14.6 and 9.2 basis points respectively), but not as high as African-Americans (30.6 basis points)
• Among recent movers, only Black and Non-white Hispanics pay significantly more than White households (20.6 and 14.7 basis points)
• In the FHA/VA market minorities rates are generally not significantly greater than those of White households. Only African-American households pay significantly higher rates (14.8 basis points)
Hispanic Homeownership Seminar - June 1, 2006 68
Home Purchase Market – Other Variables
• Higher housing cost to income ratios lead to significantly higher interest rates, but there are no obvious systematic differences in this ratio across racial/ethnic groups.
• Higher levels of education are associated with lower interest rates – these effects are generally more important in conventional markets
Hispanic Homeownership Seminar - June 1, 2006 69
• Lower levels of education characterize lower income minorities particularly Hispanics.
Conventional/Home Purchase/1st Mortgage/Full Sample
Low Income Hispanic 12.3% college educated
Low Income African-American 20.9% college educated
Low Income White 28.9% college educated
Conventional/Home Purchase/1st Mortgage/Recent Mover
Low Income Hispanic 14.1% college educated
Low Income African-American 30.3% college educated
Low Income White 38.1% college educated
Hispanic Homeownership Seminar - June 1, 2006 70
• Higher Loan to Value ratios lead to higher interest rates for recent movers in the conventional market.
Low Income Hispanic 49.4% LTV>.9
Low Income African-American 44.4% LTV>.9
Low Income White 29.7% LTV>.9
High Income Hispanic 53.3% LTV>.9
High Income African-American 39.0% LTV>.9
High Income White 27.3% LTV>.9
Hispanic Homeownership Seminar - June 1, 2006 71
• African American households lower value (lower quality) units are correlated with higher interest rates.
Conventional/Home Purchase/1st Mortgage/Full Sample
Low Income Hispanic $149,954 house value
Low Income African-American $109,833 house value
Low Income White $160,217 house value
Hispanic Homeownership Seminar - June 1, 2006 72
Refinancing
• For AHS this is defined as a loan originated in a year more recent than the year of purchase
• Refinancing is less likely to be done in the FHA/VA market than Conventional (1,089 versus 5366)
• Minorities, particularly African-Americans appear somewhat less likely to refinance
Conventional/Refinance/1st Mortgage/Full Sample
Hispanic 22.8% refinance*
African-American 22.5% refinance*
White 29.7% refinance**percentages = refinancings as a % of to first mortgages
Hispanic Homeownership Seminar - June 1, 2006 73
Regression Results
• Interest Rate Differential by Racial/Ethnic Group
• For FHA/VA loans no significant differential exists
• For Conventional loans, blacks and non-white Hispanics refinance at a significantly higher cost (41.2 and 12.8 basis points respectively).
Hispanic Homeownership Seminar - June 1, 2006 74
Other Variables Impact on Interest Rates
• College Educated Households have significantly lower rates, but generally education less important in the refinancing market
• Generally risk related control variables less important in the FHA/VA market than the conventional
• In particular, housing cost to income ratios, loan terms, and house value all behave as expected.
• Lower house values again lead are associated with significantly higher rates for African-American households
Hispanic Homeownership Seminar - June 1, 2006 75
Junior Mortgages and Home Equity Loans
• Minorities are much less likely than whites to use home equity loans, more comparable to whites in use of junior mortgages
Low Income Hispanic 4.1% home equity*
Low Income African-American 3.8% home equity*
Low Income White 12.3% home equity**home equity loans / total first mortgage loans
Low Income Hispanic 5.2% junior mortgage*
Low Income African-American 7.6% junior mortgage*
Low Income White 8.9% junior mortgage**junior mortgages / total first mortgage loans
Hispanic Homeownership Seminar - June 1, 2006 76
• Generally Hispanic Interest Rates are comparable to Whites while African-Americans rates are substantially higher
Junior Mortgage
Low Income Hispanic 8.13% interest rate
Low Income African-American 8.89% interest rate
Low Income White 8.19% interest rate
Home Equity Loans
Low Income Hispanic 6.53% interest rate
Low Income African-American 7.74% interest rate
Low Income White 6.68% interest rate
Hispanic Homeownership Seminar - June 1, 2006 77
• Whites and Hispanics take on substantially more debt than African-American families
Junior Mortgage
Low Income Hispanic $37,591 amount of debt
Low Income African-American $21,749 amount of debt
Low Income White $34,512 amount of debt
Home Equity Loans
Low Income Hispanic $26,142 amount of debt
Low Income African-American $21,399 amount of debt
Low Income White $22,916 amount of debt
Hispanic Homeownership Seminar - June 1, 2006 78
Regression Results
• Significance of Race/Ethnic Group
– For junior mortgages only African-Americans have a significantly higher interest rate controlling for other factors (44.7 basis points higher)
– For home equity loans, Non-white Hispanics and African-Americans have significantly higher interest rates controlling for other factors (52.3 and 62.7 basis points respectively)
Hispanic Homeownership Seminar - June 1, 2006 79
Other Significant Variables
Junior Mortgages
• Education level is inversely correlated with interest rates and whites are generally better educated than minorities.
• Higher housing Cost to income ratios, longer loan terms and lower house values are consistent with higher interest rates. African-Americans have substantially lower house values than other households.
Hispanic Homeownership Seminar - June 1, 2006 80
Other Significant Variables
Home Equity Loans
• High housing cost to income and lower house values give rise to significantly higher interest rates.
• African-Americans are observed to have relatively high interest cost, in part, because of their relatively low property values.
• Hispanics interest rates are higher than they otherwise would be because of their high levels of debt (primary component of housing cost) relative to their income level.
Hispanic Homeownership Seminar - June 1, 2006 81
Summary
• First look at large sample of Hispanic families, first look at junior and secondary financing.
• Controlling for other factors that might influence rates, minority households are more likely to have higher rates for 1st mortgages than whites, but primarily in the conventional markets.
• African-Americans rates tend to higher than Hispanics or whites in the refinancing market.
• African-American and Hispanic households are observed to have lower levels of educational attainment which are correlated with higher interest rates particularly in the conventional market.
Hispanic Homeownership Seminar - June 1, 2006 82
Summary
• African-Americans generally appear to be in lower quality housing units which are correlated with higher interest rates.
• Hispanics tend to have relatively high loan to value ratios on their 1st mortgages compared to other households (particularly whites)
• Hispanics and African-Americans are slightly less likely to use junior mortgages and substantially less likely to use home equity loans than white households.
• Hispanics have relatively high debt levels when they do use secondary financing, but African-Americans interest rates tend to be substantially higher than the other two groups in this market.