Series: own p PP statistic

Determinants of Long-Run Homeownership
Rates: Evidence from Taiwan
CHIEN-WEN PENG
N AT I O N A L TA I P E I U N I V E R S I T Y
I-CHUN TSAI
N AT I O N A L U N I V E R S I T Y O F
KAOHSIUNG
STEVEN BOURASSA
UNIVERSITY OF LOUISVILLE
06/25/ 2010
Homeownership Rate
Number of Owner - occupied Houinsg Units
Number of Household s
Accumulated results of individual
household’s housing tenure choice.
Benefits of Homeownership
 Positive impacts on people’s behavior, especially
during the childhood. (Green and White 1997;
Haurin et al. 2002; Lien et al. 2008)
 higher test scores
 Increase people’s attachment to their property and
community, which tends to have stabilizing effect
on society. (Rossi and Weber 1996; Dipasquale
and Glaeser 1999)
better neighbor, better citizen
Policies to Promote Homeownership Rate

Supply Side Subsidy Affordable Public Housing

Demand Side Subsidy
 Preferential Interest Mortgage
 Mortgage Interest Deduction from Income Tax
 Lower Property Tax Rate
 Lower down payment Required (Higher LTV)
Costs of Homeownership
Obscure costs with respect to
 Limited economic resource allocation
 Economic development
 Housing market operation
Homeownership Rates in US-1965~2008
80
70
60
50
67.5%
63.4%
+4.1%
40
30
20
10
0
1965
1970
1975
1980
1985
1990
1995
2000
2006
Case & Shiller House Price Index-1987~2009
210
190
170
150
130
110
90
70
50
Index
Annual Change
+206.2%
189.93
132.64
62.03
-30.16%
1987Q1 1989Q1 1991Q1 1993Q1 1995Q1 1997Q1 1999Q1 2001Q1 2003Q1 2005Q1 2007Q1 2009Q1
20%
15%
10%
5%
0%
-5%
-10%
-15%
-20%
-25%
House Price and Homeownership Rate

House Price
 Relative Cost of Owning vs. Renting
 House Price Affordability (wealth and income
constrains)

House price
↑ User Cost of Owning ↑
Affordability ↓
↓
Exp. House Price Appreciation↑ Ownership Rate↑
Ownership Rate
Homeownership Rates and House Price in US
70
200
69
180
68
160
67
140
Positive or Negative?
66
Ownership
65
HPI
64
120
100
80
60
40
62
20
61
0
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
03
20
04
20
05
20
06
20
07
20
08
20
09
63
Painter and Redfearn(2002)
Interest rates had an influence on both
housing supply and timing of changes of
tenure status from renter to owner, the longterm homeownership rate appears
independent of interest rates.
To promote homeownership rates, low down
payment and improved technology for
assessment of credit risk may be more
effective.
Homeownership Rates in Taiwan:1976~2008
87.4%
%
90
+20%
85
80
75
70
65
60
67.4%
1976
1981
1986
1991
1996
2001
2006
Ownership Rates in Taiwan and USA-1976~2008
%
87.4%
90
+20%
85
Taiwan
USA
80
75
67.4%
70
+2.7%
65
64.8%
67.5%
60
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Ownership Rates and House Price of Taipei City
Taipei City
p
own
30
85%
80%
25
75%
20
70%
15
65%
10
60%
20
06
20
04
20
02
20
00
19
98
19
96
19
94
19
92
19
90
19
88
19
86
50%
19
84
0
19
82
55%
19
80
5
19
80
19
81
19
82
19
83
19
84
19
85
19
86
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
Ownership Rates and House Price of Taipei County
Taipei County
p
own
16
95%
14
90%
12
10
6
4
85%
80%
75%
8
70%
65%
60%
2
55%
0
50%
Ownership Rates and House Price of Taichung City
Taichung City
p
own
14
90%
12
85%
80%
10
75%
8
70%
6
65%
4
60%
20
06
20
04
20
02
20
00
19
98
19
96
19
94
19
92
19
90
19
88
19
86
50%
19
84
0
19
82
55%
19
80
2
19
80
19
81
19
82
19
83
19
84
19
85
19
86
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
Ownership Rates and House Price of Kaohsiung City
Kaohsiung City
14
p
own
12
10
4
2
0
95%
90%
85%
80%
8
75%
6
70%
65%
60%
55%
50%
Research Questions
 Both of the patterns of long-run
homeownership rates and house prices in
US. and Taiwan are strange.
 What are the determinants of long-run
homeownership rates? (Does it implies
Taiwan’s homeownership promotion policies
are more effective than U.S.? )
Literature Review
 Abundant Literature on Determinants of
Individual Household’s Tenure Choice
 Some studies focus on Homeownership
Rates Differences in different Nations
/Regions
 Rare on the Determinants of Long- run
Homeownership Rates
Tenure Choice- Market Factors
 Housing Price, HP Fluctuation Risk ↑ Rent
 Borrowing Constrains (LTV↓, Interest Rate↑) ↑
Rent
 Rent, Rent Fluctuation Risk ↑ Buy
 Expected Housing Price Appreciation ↑Buy
Tenure Choice-Institution Factors
Property Tax ↑ Rent
Relative Cost of Owning vs. Renting ↑ Rent
Deduction of Mortgage Interest from Income
Tax↑ Buy
Owner-occupied Housing Subsidies↑ Buy
Tenure Choice- Household’s Characteristics
Expected Mobility ↑ Rent
Household Income↑ Buy
Household Head’s Age↑, Married  Buy
Family Size ↑ Buy
Number of Dependent Children↑ Buy
Selected Variables (no institutional factors )
 House Price (p)
 Household Income (I)
 House Price to Income Ratio (pI)
 Rent Growth Rate (red)
 House Price Growth Rate (pd)
 Income Growth Rate (ld)
 Household Growth (h)
 Mobility Rates (mov)
 Proportion of Married Couples (mar)
 Proportion of Elderly People (old)
Empirical Study
 Investigate the Determinants of Long-Run
Homeownership Rates
 Data: Taipei City, Taipei County, Taichung City,
Kaohsiung City, 1980~2007,Sample Size 112
 Methodology: Panel Co-integration
Panel Co-integration
 Cointegration is an econometric
property of time series variables.
 If two or more series are themselves
non-stationary, but a linear combination
of them is stationary, then the series
are said to be cointegrated.
 Panel Co-integration= Cross Section + Time
Series  More Samples, More Information
Panel Unit Root Test


IPS
ADF-Fisher
Variable
Panel Unit Root Test
IPS
ADF - Fisher Chi-square
Levels
own
mar
mov
old
h
p
I
pI
pd
Id
red
0.27
7.08
-1.35
6.55
-1.58
0.01
-0.10
-1.19
-2.00 **
-4.92 ***
-0.67
9.62
0.12
21.53 ***
0.19
13.32
6.03
5.56
10.75
17.70 **
38.68 ***
8.22
Variable
Differences
△own
△mar
△mov
△old
△h
△p
△I
△pI
△pd
△Id
△red
Panel Unit Root Test
IPS
-13.23
-6.09
-8.49
-6.37
-9.62
-2.05
-5.71
-5.70
-11.70
-7.38
-4.22
***
***
***
***
***
**
***
***
***
***
***
ADF - Fisher Chi-square
105.03
48.18
69.03
49.81
77.28
17.65
45.45
44.67
88.14
61.41
32.36
***
***
***
***
***
**
***
***
***
***
***
Results of Panel Unit Root Test
 Can not reject the null hypothesis of having
a unit root for the levels of most variables,
except house price appreciation rate (pd)
and income growth rate (Id).
 The differences of all variables are
significantly to reject the null hypothesis
which implies most variables are I(1).
Panel Co-integration Test
own and mar mov old h (demographic)
own and I, p, pI (affordability)
own and red (consumption)
 Model 1 without trend
 Model 2 with trend
Panel Statistics
Weighted
Panel Statistics
Group Statistics
Series: own mar mov old h
PP statistic
-4.605 ***
-4.605 ***
-6.181
ADF statistic
-4.519 ***
-4.519 ***
-6.048
PP statistic
-2.875 ***
-3.336 ***
-3.652
ADF statistic
-2.499 **
-3.031 ***
-3.334
PP statistic
-3.161 ***
-3.339 ***
-3.504
ADF statistic
-3.118 ***
-3.295 ***
-3.455
PP statistic
-4.875 ***
-4.821 ***
-5.551
ADF statistic
-5.269 ***
-5.416 ***
-5.418
PP statistic
-0.436
-0.658
-0.065
ADF statistic
-0.385
-0.726
0.084
Series: own mar
Series: own mov
Series: own old
Series: own h
Weighted
Panel Statistics
Group Statistics
Panel Statistics
Series: own p I
PP statistic
-3.563 ***
-2.823 ***
-3.248 ***
ADF statistic
-4.160 ***
-3.856 ***
-4.338 ***
PP statistic
-1.723
-1.919
-1.054
ADF statistic
-1.674
-1.860
-1.043
PP statistic
-4.203 ***
-2.986 ***
-3.736 ***
ADF statistic
-3.644 ***
-3.089 ***
-3.247 ***
Series: own p
Series: own I
Series: own pI
PP statistic
0.919
1.246
2.053 **
ADF statistic
1.195
1.305
2.190 **
Panel
Statistics
Series: own red
PP statistic
-0.421
ADF
statistic
-0.523
Weighted
Panel
Statistics
Group
Statistics
-0.216
0.464
-0.314
0.361
Panel Co-integration Test without trend
 Long-run equilibrium relationship between
own and I, mar, old, mov
 No cointegration relationship between own
and h, p, red
Panel Co-integration Test -With Trend
 own and mar mov old h (demographic)
 own and I, p, pI
 own and red
(affordability)
(consumption)
Panel Statistics
Weighted
Panel Statistics
Group Statistics
Series: own mar mov old h
PP statistic
-6.77
***
-7.02
***
-8.20
***
ADF statistic
-6.71
***
-6.70
***
-6.39
***
PP statistic
-5.95
***
-5.99
***
-5.76
***
ADF statistic
-5.93
***
-5.97
***
-5.72
***
PP statistic
-5.30
***
-5.12
***
-4.94
***
ADF statistic
-5.28
***
-5.12
***
-5.00
***
PP statistic
-6.92
***
-6.17
***
-6.52
***
ADF statistic
-6.92
***
-6.17
***
-6.48
***
PP statistic
-5.47
***
-4.43
***
-4.98
***
ADF statistic
-5.49
***
-4.49
***
-5.05
***
Series: own mar
Series: own mov
Series: own old
Series: own h
Panel Statistics
Weighted
Panel Statistics
Group Statistics
Series: own p I
PP statistic
-7.51
***
-6.76
***
-8.65
***
ADF statistic
-7.33
***
-6.61
***
-7.41
***
PP statistic
-8.08
***
-8.91
***
-7.82
***
ADF statistic
-8.03
***
-8.76
***
-7.47
***
PP statistic
-7.08
***
-5.19
***
-6.88
***
ADF statistic
-7.93
***
-6.38
***
-6.68
***
PP statistic
-5.63
***
-4.82
***
-5.31
***
ADF statistic
-5.61
***
-4.83
***
-5.39
***
Series: own p
Series: own I
Series: own pI
Weighted
Panel Statistics
Panel Statistics
Group
Statistics
Series: own red
PP statistic
ADF statistic
-7.28
-7.29
***
***
-6.79
-6.79
***
***
-6.55
-6.57
***
***
Panel Co-integration Test with trend
 All variables have cointegration
relationships with homeownership rates.
 A trend in homeownership rate serial.
FMOLS_ Taipei City
variable
coefficient
t value
MAR
2.41
7.19
MOV
0.25
1.30
OLD
4.02
8.15
H
-0.09
-0.33
P
0.16
1.81
I
0.01
0.56
RED
0.07
1.09
FMOLS_ Taipei County
variable
coefficient
t value
MAR
-0.23
-1.19
MOV
0.28
2.48
OLD
4.77
5.36
H
-0.61
-2.77
P
0.22
1.47
I
-0.12
-2.90
RED
-0.07
-1.02
FMOLS_ Taichung City
variable
coefficient
t value
MAR
-0.44
-1.10
MOV
-0.31
-1.06
OLD
-0.42
-0.24
H
-0.44
-1.31
P
0.49
1.04
I
0.16
3.84
RED
-0.02
-0.08
FMOLS_ Kaohsiung
variable
coefficient
t value
MAR
0.77
0.23
MOV
-0.07
-0.18
OLD
2.76
0.55
H
0.25
0.36
P
-0.04
-0.09
I
0.22
2.06
RED
0.24
0.97
FMOLS_ Panel
variable
coefficient
t value
MAR
0.63
2.57
MOV
0.04
1.27
OLD
2.78
6.90
H
-0.22
-2.02
P
0.21
2.12
I
0.07
1.78
RED
0.06
0.48
Results of FMOLS
•the most influential variables of own
are different in the four cities.
•Taipei City: old(+), mar(+), p(+)
•Taipei County: old(+), mov(+), l(-), h(-)
•Taichung City and Kaohsiung City: I (+)
•In General, old, mar, p, I (+) , h (-)
Conclusions
 A trend exists in Taiwan’s homeownership rates,
not explainable by selected variables which may
contributed to the influence of institutional factors.
 If not consider the trend, long-run equilibrium
relationships only between ownership rates and
 household income
 proportion share of married couples
 Proportion of elderly people
 mobility rates
Conclusions
 If consider the trend, can find co-integration
between homeownership rates and house prices,
household growth rate, rent growth rate.
 From FMOLS,
 the most influential variables of own are different
in the four cities.
 In general, proportion of elderly people, proportion
of married couple, house price are most influential
vars.
Policies Implications
Why there is a trend in Taiwan’s
homeownership rates?
Possible explanation:
 Low owning cost which due to low property
tax and high expectation of house price
appreciation, especially in Taipei City
 effective property tax rate↑
 better rental housing market
Thanks for your Attention