Transcript Empirical Evidence - ERES Digital Library

Sale of Public Housings to SittingTenants: What Determines the Success
or Failure of Such Privatization
Programs around the World since
1980's ?
Kwok Chun, Wong and Linzi, Zheng
University of Hong Kong
Focus of this presentation
• Drawing a general pattern of public housing
production proceeds for different countries
can by examining Chinese indemnificatory
housing construction, try to explain why the
outcomes of this program have not been close
to satisfactory every since 1994.
Introduction
• Comparisons and analysis of indemnificatory
housing agents and responsibilities in different
countries
• Hypothesis of China’s failure and other
countries’ successes
• Model construction within principal—agent
theory
• Empirical Evidences
Comparisons and analysis
Table 1 indemnificatory housing provider and their responsibilities
Agent
Singapore
HUB in 1966
Responsibilities
HUB is responsible for providing houses for middle and low income
families according the central government’s policy.
Draw housing policy and develop implementation system; making
Japan
Province House Bureau in budgets and allocate nation housing resources; drawing national five1949
years' program; monitoring housing relate activities of local
government, public group, relate financial department and etc.。
Comparisons and analysis
• In Japan and Singapore, central governments, local housing
management departments and construction agencies are
participants of policy making
• central governments are premier whose responsibility is
coordination of related resources
• local housing management and construction departments are
policy implementers whose responsibilities are clearly
regulated by policy
• The local governments do not have rights to implementation
like China does
Comparisons and analysis
• In China, a series of indemnificatory policies have been
enacted out ever since 1994
• local governments are they actors of indemnificatory housing
program
• However, there are no specific codes to regulate local
government’s relative behaviors
• Local governments have incentives to act tangentially to
central government’s objectives.
Hypothesis
• relative successes of countries like Japan and
Singapore is because of their adoption of
principals and agents and the information
transportation chain between them
• central governments and local governments
actually act as principal and agents constraint
by contract and local governments can
achieve their maximum utility by providing
minimum units of indemnificatory housing.
Model construction—Assumptions
• local governments have enough information advantage to
make the central government lack of real information because
the transportation chain is too long
• the central government wants to maximize social welfare as a
whole while local governments aim to achieve GDP growth
rate as high as possible
• central government is risk neutrality for the enactments are
implemented within the whole country while local
governments are risk aversion when certainty equivalents and
risk premium should be considered.


Model construction
• a a  A,
• y means production, and y = a+ is the production
function of local governments
• W（y） = s + by , b [0 ， 1 ]
• y = y – W denotes income of central government
• The problem the central government is facing is to
maximize W and rewards and punishments according
to observable y


Model construction
•
•
•
•
Risk aversion of local governments, U(A)= -e  rA
'
' '
local governments are C( a), C(0)=0, C (a)>0, C (a)>0
2
ka
let C（a）= / 2 , k > 0 means cost coefficient
Therefore, the real income of local governments are:
 = W - C( a) = s +b( a +  ) - ka / 2
2
（1）


Model construction
• let X equals to the stable income of local
governments and Y is random income, if u (X ) = Eu (
Y), we call X is Y’s certainty equivalence, so we know
the certainty equivalence of local governments is：
CE(  ) = E(  ) - R, R>0
（2）
Model construction
• It is easy to prove that R = rb 2  2 /2 and because that
2
there is  = s +b( a +  ) - ka / 2, we get:
CE(  ) = s + ba - ka /2 - rb  2 /2
2
2
（3）
• Let means the reserve income level of local
governments. Thus the constraint of local
governments’ participation of this program is:
( PC)
s + ba - ka /2 - rb  2 /2  
2
2
（4）
Model construction
• The agent decision is:
2
2
2
max CE(  ) = s + ba - ka /2 - rb  /2
（5）
*
• The incentive consistent condition is a = b/k
• According to the given contract function, W (y)= s + by,
the principal decision is：
'
maxE(y = y – W)
（6）
=( - s + (1 - b) a)
2
2
2
s.t. s + ba - ka /2 - rb  /2   ( PC)
*
a = b/k
( IC)
Model construction
• by solving these equations we can get the optimal
incentive coefficient of central government is:
1
1  rk 2
1
b* =
1  rk 2 into
b* =
• Substitute
optimal effort degree is:
*
a =
1
k (1  rk 2 )
*
a = b/k we can get the
Process of calculation
Measurement of explicit cost
• The Explicit Cost, includes land grant fee, housing
preferential favorable fee and construction cost of
infrastructure.
• Opportunity Cost. The potential decrement of GDP
caused by reducing investments in fixed assets.
Process of calculation
•
land grant fee is the payment to land which should be repaid by the seller
when owner wants to resell one’s economic apartment five years after
buying which is about 10% of the final price of this
Table 2 land grant fee of Beijing
Average
price
commercial housing
of
Sales
of
economic
housing
（ Yuan/square meters ） （million square meters）
Trade volume
（million）
1999
4787
45.8
219244.6
2000
4557
168.2
766487.4
2001
4716
185.3
873874.8
2002
4467
220.7
985866.9
2003
4456
320
1425920
2004
4747
306.3
1454006
2005
5853
222.87
1304458
Total repayment to land
(million)
Average repayment to
land（million）
702985.8
100426.5
Process of calculation
Table 3 housing preferential favorable fee
item
1
Original standard
Implementation standard
Comprehensive development of urban construction management One million Yuan each item, five million Yuan for un-comprehensive items and two million Yuan for Cancelled
fees
others
2
Ownership registering fee
80Yuan per unit
Cancelled
3
Land acquisition and management fees
1.5％ of total land acquisition costs
Cancelled
4
Temporary land use and management fees
Cancelled
Cancelled
5
Fees of urban house pulling down and accommodation
0.3％ of pulling down and accommodation
Cancelled
6
Transaction fees
1％of housing price
Cancelled
7
Housing safety evaluation fees
0.8－1.2Yuan /sq.m
Cancelled
8
Fire safety licensing
10元
Cancelled
9
Infrastructure construction cost
Residential 160Yuan/sqm
80Yuan/sqm
Non-residential 200 Yuan/sqm
100Yuan/sqm for non-residential housing.
10
City Fund for energy saving and the development of new wall 8Yuan/sqm
for
residential
housing
and
4 Yuan/sqm
materials
11
Flood control costs
20 acreage
10 acreage
12
Construction license licence fees
0.1％－0.3％of investment
0.05％－0.15％of investment
13
Road charges
0.4－3 Yuan/day
0.2－1.5Yuan/day
14
Greening payments
240（suburban）
3000－5000Yuan when greening percentage is less
145（outer suburbs）
than 30%
15
loan note
120 Yuan/note
60 Yuan/note
16
Market transaction service fee
0.1％ of the price of tender bid
0.05％of the price of tender bid
Process of calculation
Table4 Total explicit cost of economic housing each year in Beijing
housing
construction
preferential
cost
favorable fee
Land grant fee
Total explicit cost
of
economic
housing each year
(million)
70298.58
20985.71
infrastructure
53513.57
of
Total
142787.9
Process of calculation
Measurement of opportunity cost
• Let GDP = a + b* fixed assets investment
fixed
Year
GDP（Billion）
1990
500.8
135.6
1991
598.9
144.4
1992
709.1
201
1993
886.2
318.2
1994
1145.3
507.9
1995
1507.7
841.5
1996
1789.2
876.9
1997
2075.6
961.2
1998
2376.0
1155.6
1999
2677.6
1170.6
2000
3161.0
1297.4
2001
3710.5
1530.5
2002
4330.4
1814.3
2003
5023.8
2517.1
2004
6060.3
2528.3
2005
6886.3
2827.2
assets
（Billion）
investment
Process of calculation
• Let GDP = a + b* fixed assets investment
GDP = 55.34 + 2.26 fixed assets investment
（t＝12.3497）
2
R ＝0.975 F＝587.446 n＝15
Process of calculation
Table 6 opportunity cost measured by GDP
GDP（billion）
1999
2677.6
2000
3161
2001
3710.5
2002
4330.4
2003
5023.8
2004
6060.3
2005
6886.3
GDP loss（％）
1.22%
1.04%
0.88%
0.76%
0.65%
0.54%
0.48%
Incentive level analysis
Table 7 the proportion of cost distribution
Cost
of
local
Cost
of
central
The
responsible
government
government
percentage
of
（million）
（million）
government(%)
1999
31606.12
6577.388
82.77%
2000
113365.1
22994.67
83.14%
2001
126952.7
26216.29
82.88%
2002
147359.2
29576.06
83.28%
2003
213414.4
42777.65
83.30%
2004
210516.9
43620.23
82.84%
2005
170430.9
39133.79
81.33%
local
Risk aversion degree analysis of
local governments
According to results of the former two parts we can get
conclusions twofold:
One is the cost of efforts of local governments in China is
very high and the incentive level offered by central
government is low, therefore the risk aversion degree of
local governments is very high.
Empirical Evidence
Assumptions
1）because of the impossible and very expensive
monitoring cost of local governments, the central
government choose to be accept the core role of local
governments within the utility function.
2）GDP and indemnificatory housing have negative
relationship.
Empirical Evidence
Utility function of central government
U x 2 = F (TGDP，H)
where H is the amount of indemnificatory housing.
Suppose central government has the minimum requirement of local
2
2
2

s
+
ba
ka
/2
rb
/2  
indemnificatory housing production according to
we get H  minH, where minH satisfies s + ba - ka 2 /2 - rb 2  2 /2 = .
Empirical Evidence
utility function of local government
UX 1 = F (UX 2 ， LGDP )
= F (f ( TGDP， H ) ， LGDP)
U x 2
> 0,
H
u x1
F U x 2
F
LGDP
=
+
LGDP H
H U X 2 H
F
Obviously U
X2
> 0， U x2 > 0
H
(*)
so the first item of eq(*) is positive and the
second one is negative. When it is decreasing function, within the
principal agent relationship, local governments are inclined to keep H
above the stable value.
Empirical Evidence
utility function of local government
maxUX 1 = F ( f ( TGDP，H ) ，LGDP)
s. t. H  minH
H， LGD P  0
Build a lagrange function：
Z = F ( f ( TGDP，H ) ，LGDP) +  (H- min H)
Has to satisfy
Z F
=
+0
H H
Z
= H - MinH

H  0 and H
  0 and 
Z
=0

Z
=0
H
(**)
Empirical Evidence
utility function of local government
In eq(**), because H>min H> 0,
Z
F
= 0 and
+  ＝0 , From eq(*) we know
H
H
Z
F
F
< 0，so  ＝－
= 0，
 0 and

H
H
therefore we get H =min H, which means only the minimum
amount of indemnificatory housing provided by local governments
can satisfy the requirements of utility maximization.
Empirical Evidence
Table 8 middle and low income groups in Beijing
Year
Total residents（million）
1999
Middle
and
low
income
Middle and low income families
individuals（million）
（million）
1099.8
329.9
103.1
2000
1107.5
332.3
103.8
2001
1122.3
336.7
105.2
2002
1136.3
340.9
106.5
2003
1148.8
344.6
107.7
2004
1162.9
348.9
109.0
2005
1180.7
354.2
110.7
Empirical Evidence
Indemnificatory housing demand and level in Beijing
Middle
Year
and
income
low
families
Amount
of
indemnificatory
Indemnificatory
housing
Indemnificatory
housing in need（ten thousand
completed
（million）
units）
（ten thousand units）
1999
103.11
30.93
1.29
4.17%
2000
103.81
29.86
1.65
5.55%
2001
105.22
29.84
2.16
7.26%
2002
106.53
29.67
2.05
6.94%
2003
107.70
30.09
2.70
9.24%
2004
109.02
29.72
2.73
9.22%
2005
110.69
30.19
2.94
9.74%
level
Empirical Evidence
• All the indemnificatory levels are lower than ten percent
which is obviously not sufficient. However, during this period
the GDP growth rate in Beijing has been kept on 10% --it is
good performance of economy. The GDP growth rate is stable
which means construction the indemnificatory housing
construction did not encroach on economy so we can
conclude that the minH of Beijing is some figure between 20
thousand to 30 thousand which is consist to the reality—
according the publicized political goal.
Conclusion
In China, the process of indemnificatory housing production is actually a
contract between central government and local governments, the two of
which comprise an relationship of principal and agent. Because of the
extremely high or even impossible monitoring cost and asymmetry
information, the utility functions of local governments get the maximum
point when they provide minimum amount of indemnificatory housing.