Transcript The Determinants and Long-Term Projections of Saving Rates in

The Determinants and Long-Term Projections
of Saving Rates in Developing Asia
Charles Yuji Horioka
(Institute of Social and Economic Research, Osaka University, and
National Bureau of Economic Research)
and
Akiko Terada-Hagiwara
(Asian Development Bank)
Prepared for presentation at the Federation of ASEAN Economic
Associations (FAEA) conference, Manila, Philippines, on
November 28, 2012
1
I. INTRODUCTION
2
Introduction (1)
Developing Asia has been characterized by
high saving rates almost across the board in
recent years, which have made possible high
levels of domestic investment but have also
led to large capital outflows (current account
surpluses) (see, for example, the data
presented in Kim and Shin (2009)).
3
Introduction (2)
To put it another way, the developing
economies of Asia have oversaved and
underinvested, leading to large current
account imbalances (surpluses), as asserted by
Ben Bernanke and others.
4
Introduction (3)
However, some assert that because population aging
will occur at a rapid rate in Asia, there will be a
sharp decline in saving rates in Asia. If so, the
large current account imbalances (surpluses) that
Asia is currently showing will go away by
themselves without any need for government
intervention. Is this assertion correct???
5
The Purpose of This Presentation
(1) To present data on domestic saving rates in
developing Asia during the 1966-2007 period
(2) To analyze the determinants of domestic saving
rates in developing Asia during the 1966-2007
period
(3) To project trends in domestic saving rates in
developing Asia for the 2011-30 period based
on our estimation results
6
The Sample of Countries
The following 12 Asian countries:
People’s Republic of China; Hong Kong, China; India;
Indonesia; Republic of Korea; Malaysia;
Pakistan; Philippines; Singapore; Taiwan;
Thailand, and Viet Nam
These countries comprise 95% of the GDP of
developing Asia.
7
II. DATA ON DOMESTIC SAVING
RATES IN DEVELOPING ASIA
8
India
Indonesia
Korea, Rep. of
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
Malaysia
China, People's Rep. of
Pakistan
Philippines
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
Singapore
Taipei,China
Thailand
Viet Nam
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70 75 80 85 90 95 00 07
70758085 90 95 00 07
0
20
40
60
0
20
40
60
0
20
40
60
Hong Kong, China
Nominal measure
Graphs by ctry
Real measure
Domestic Saving Rates in Emerging Asia, 2000-07
60.0
Domestic Saving Rate (percent)
50.0
40.0
30.0
20.0
10.0
0.0
PRC
HKG
INO
IND
KOR
MAL
Country
PAK
PHI
SIN
THA
TAP
VIE
The Level of Saving Rates
There have been enormous variations in the
level of saving rates, even within developing
Asia, during the past 50 years, with saving rates
tending to be higher in East Asia and Southeast
Asia (with the exception of the Indonesia,
Philippines, and Vietnam) and lower in South
Asia.
11
Trends in Saving Rates (1)
There have been enormous variations in trends
over time in saving rates, even within
developing Asia, during the past 50 years, with
saving rates showing upward trends in Korea,
Malaysia, China, Pakistan, Singapore, Taiwan,
Thailand, and Vietnam, downward trends in
Hong Kong, Indonesia and the Philippines, and
no clear trend in India.
12
Trends in Saving Rates (2)
In the 1960s, Indonesia, Korea, Pakistan, and
Singapore had among the lowest saving rates
(below 10%) with that of Singapore even being
negative, but the saving rates of these
economies increased sharply in the 1970s
except for Pakistan. By the mid-1980s, the
domestic saving rates of these economies as
well as those in Taipei, China and Malaysia
exceeded 30% of GDP.
13
III. THE DETERMINANTS OF
DOMESTIC SAVING RATES IN
DEVELOPING ASIA
14
Previous Studies (1)
• There have been many studies that have made
use of cross-section or panel cross-country
data or of time series data for individual
countries to analyze the determinants of
national saving starting with Modigliani’s
seminal 1970 paper.
15
Previous Studies (2)
Higgins, M. (1998), “Demography, National Savings, and
International Capital Flows,” International Economic
Review, vol. 39, no. 2, pp. 343-369.
Bosworth, Barry, and Chodorow-Reich, Gabriel (2007),
mimeo. (Brookings Institution)
Kim, Soyoung, and Lee, Jong-Wha (2008), Japan and
the World Economy)
Park, Donghyun, and Shin, Kwanho (2009), ADB
Economics Working Paper Series, no. 158, Asian
Development Bank
16
Previous Studies (3)
Previous studies examine the impact of such
factors as the age structure of the population,
the level and growth rate of income, life
expectancy, and the degree of financial sector
development on saving rates.
17
Determinants of Saving (Park and Shin, 2009)
All
Asia
All Countries
Countries
Minus Asia
GDP Growth
+
+
+
Lagged GDP Growth
+
Per Capita Income
+
Per Capita Income Squared
+
Life Expectancy
+
Age Dependency Ratio
(-)
Youth Dependency Ratio
Financial Development
18
Determinants of Saving (1)
• We analyze the determinants of variations
over time and among countries in developing
Asia using both random effects and country
fixed effects models.
• Following previous studies, the observations
are five-year averages except for the most
recent period (1966-70, 1971-75, 1976-80,
1981-85, 1986-90, 1991-95, 1996-2000, and
2001-07).
19
Determinants of Saving (2)
(Demographic variables)
• AGE = Aged dependency ratio (population
aged 65 and older/population aged 15-64)
• DEP = Youth dependency ratio (population
aged 0-14/population aged 15-64)
20
Determinants of Saving (3)
(GDP-related variables)
• LNGDP = The log of real per capita GDP
• LNGDPSQ = The square of LNGDP
• CHGDP = The growth rate of real per capita
GDP
21
Determinants of Saving (4)
(Financial variables)
• CREDIT = The ratio of private credit by deposit
money banks and other financial institutions to
GDP
• CREDITSQ = The square of CREDIT
• RINT = The real interest rate
22
Determinants of Saving (5)
(Other economic variables)
• FISC = The ratio of the fiscal balance of the
government to GDP
• SSR = The social security ratio (the ratio of
expenditures on social services and pensions to
gross national disposable income)
• Dropped from the final specification because not
significant
23
Descriptive Statistics
Appendix Table 2: Descriptive Statistics
Variable
No. of Obs.
Mean
Std. Dev.
Minimum
Maximum
SR
78
25.2
14
-3.8
60.2
AGE
78
7.8
2.2
5.7
16.3
DEP
78
56.8
18.6
20.3
87.6
LNGDP
78
8.5
0.9
6.7
10.5
CREDIT
78
0.7
0.5
0.1
2.2
CHGDP
78
4.6
2.8
-2.7
12.3
RINT
70
1.1
3
-7.5
8.7
Note: Refer to Appendix Table 1 for variable definitions and data sources.
24
40
60
16
20
60
DEP
80
100
6
7
8
9
LNGDP
10
11
0
.5
1
1.5
CREDIT
2
2.5
F. Social service expenditure
0
20
PWTSR
40
60
E. Per capita GDP growth
0
0
20
PWTSR
40
60
D. Log per capita GDP
40
60
14
40
10 12
AGE
20
8
0
0
6
PWTSR
C. Private credit %GDP
20
PWTSR
40
60
B. Youth dependency ratio
20
PWTSR
40
20
0
PWTSR
60
A. Aged dependency ratio
-5
0
5
CHGDP
10
15
0
5
10
15
SSR
25
Estimation Results
26
Estimation Results without Lags
Model
1
2
3
4
5
6
AGE
-0.95 **
0.35
-2.71
-0.98 **
0.38
-2.58
-0.83 **
0.36
-2.28
-1.58 ***
0.52
-3.03
-1.63 ***
0.54
-3.02
-1.69 **
0.66
-2.55
Table 1: The Determinants of the Real Domestic Saving Rate in Developing Asia
DEP
LNGDP
LNGDPSQ
CREDIT
CREDITSQ
CHGDP
RINT
Country fixed effects model with no lags
-0.03
-43.13 ***
2.92 ***
14.48 **
-6.46 ***
0.13
4.94
0.32
4.92
1.62
-0.23
-8.73
9.24
2.94
-3.99
-0.03
-44.21 ***
2.99 ***
14.58 **
-6.41 ***
0.08
0.13
6.19
0.39
5.01
1.70
0.14
-0.22
-7.15
7.75
2.91
-3.78
0.59
0.05
-35.63 **
2.53 ***
14.88 ***
-6.25 ***
0.16
0.03
0.18
13.38
0.74
4.80
1.56
0.14
0.16
0.27
-2.66
3.40
3.10
-4.00
1.15
0.19
Random effects model with no lags
-0.08
-46.79 ***
3.15 ***
15.35 ***
-6.71 ***
0.10
8.16
0.50
5.87
1.82
-0.72
-5.73
6.29
2.61
-3.69
-0.07
-49.45 ***
3.31 ***
15.08 **
-6.39 ***
0.24 *
0.10
8.97
0.55
5.92
1.83
0.13
-0.69
-5.51
5.99
2.55
-3.49
1.75
-0.62
-40.65 **
2.85 ***
14.63 **
-6.15 ***
0.31 ** -0.04
0.15
16.77
1.01
6.23
1.84
0.15
0.19
-0.41
-2.42
2.81
2.35
-3.35
2.09
-0.21
(continued)
R-squared # obs.
0.76
0.61
0.70
0.76
0.62
0.71
0.69
0.65
0.72
78
0.75
0.68
0.74
0.75
0.70
0.75
0.66
0.75
0.78
78
78
70
78
70
27
Estimation Results with Lags
Table 1 (continued)
Model AGE
R-squared # obs.
RINT
CHGDP
CREDITSQ
CREDIT
LNGDPSQ
LNGDP
DEP
Country fixed effects model with lags of LNGDP, LNGDPSQ, CREDIT, CREDITSQ, and CHGDP
66
0.60
-3.51
6.70
1.96 ***
-24.63 **
-0.02
7 -1.34 ***
0.59
2.13
6.22
0.47
8.40
0.20
0.40
0.66
-1.65
1.08
4.17
-2.93
-0.09
-3.33
66
0.60
0.19
-3.32
6.75
2.03 ***
-25.64 **
-0.02
8 -1.41 ***
0.62
0.15
2.24
6.45
0.52
8.86
0.20
0.46
0.68
1.28
-1.48
1.05
3.90
-2.89
-0.09
-3.10
62
0.55
-0.07
0.26 *
-2.98
7.12
1.43
-14.12
0.09
9 -1.42 **
0.57
0.26
0.14
1.95
5.70
0.86
16.90
0.23
0.51
0.64
-0.27
1.84
-1.52
1.25
1.66
-0.84
0.37
-2.77
Random effects model with lags of LNGDP, LNGDPSQ, CREDIT, CREDITSQ, and CHGDP
66
0.58
-4.09
7.61
2.39 ***
-32.31 ***
-0.14
10 -1.91 ***
0.70
2.52
7.40
0.73
10.43
0.13
0.61
0.73
-1.62
1.03
3.29
-3.10
-1.08
-3.12
66
0.58
0.40 **
-3.51
7.18
2.51 ***
-34.12 ***
-0.14
11 -2.02 ***
0.74
0.18
2.56
7.63
0.83
12.04
0.14
0.68
0.76
2.17
-1.37
0.94
3.03
-2.83
-0.98
-2.99
62
0.52
-0.17
0.37 **
-2.70
5.02
1.97 *
-23.92
-0.10
12 -1.93 ***
0.73
0.24
0.17
2.40
7.37
1.16
18.59
0.16
0.72
0.76
-0.71
2.22
-1.13
0.68
1.69
-1.29
-0.61
-2.68
Notes: The figures are the estimated coefficients (first row), robust standard errors (second row), and t-values (third row). The first R-squared is
within, the second is between, and the third is overall. The constant terms and country fixed effects are not shown to save space.
*Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level.
28
Estimation Results (1)
(Demographic variables)
• AGE = Aged dependency ratio (negative, as
expected, and significant)
• DEP = Youth dependency ratio (totally
insignificant)
29
Estimation Results (2)
(GDP-related variables)
• LNGDP = The log of real per capita GDP
(negative and significant, as expected)
• LNGDPSQ = The square of LNGDP (positive
and significant)
• CHGDP = The growth rate of real per capita
GDP (positive and significant about half the
time)
30
Estimation Results (3)
(GDP-related variables) (cont’d)
Thus, LNGDP has a nonlinear (convex) impact on the
domestic saving rate, with its impact being negative
at low levels of LNGDP and positive at high levels of
LNGDP. Our results imply that, by 1976-80, income
levels had become high enough for LNGDP to have a
positive impact on the domestic saving rate in nine
out of the twelve economies in our sample (with the
only exceptions being India, China, and Viet Nam).
31
Estimation Results (4)
(Financial variables)
• CREDIT = The ratio of private credit to GDP
(positive and significant without lags)
• CREDITSQ = The square of CREDIT (negative and
significant without lags)
• RINT = The real interest rate (never significant)
32
Estimation Results (5)
(Financial variables) (cont’d)
Thus, CREDIT has a nonlinear (concave) impact on the
domestic saving rate, with its impact being positive
at low levels of CREDIT and negative at high levels of
CREDIT. Our results imply that financial sector
development has progressed enough in six of the
twelve economies in our sample (China; Hong Kong,
China; Republic of Korea; Malaysia; Singapore; and
Taiwan) for the availability of private credit to have a
negative impact on the domestic saving rate in these
economies.
33
Estimation Results (Summary)
Domestic saving rates in developing Asia are determined
primarily by the age structure of the population
(especially the aged dependency ratio), the level of
per capita GDP, and the degree of financial sector
development, with the latter two having a non-linear
impact thereon (convex and concave, respectively).
34
IV. FUTURE TRENDS IN DOMESTIC
SAVING RATES IN DEVELOPING
ASIA
35
Projections of Future Trends (1)
• We project future trends in domestic saving rates for
2011-20 and 2021-30 using:
• Our estimation results based on our country fixed
effects model
• U.N. projections of future population by age
• Asian Development Bank projections of future per
capita GDP
• Our own projections of future trends in financial
sector development based on ADB projections in
future per capita GDP in conjunction with data on
financial sector development broken down by per
capita GDP.
36
Projections of Future Trends (2)
• The United Nations projects that there will be
enormous variations in future demographic trends
among the countries of developing Asia, with the
demographic bonus ending as early as 2010 in some
countries or territories such as Hong Kong,
Singapore, and Thailand and as late as 2035-40 in
some countries such as India and the Philippines.
• Not surprisingly, we also find enormous variations in
future trends in domestic saving rates in developing
Asia.
37
Future Trends in Saving Rates
Table 2: Future Trends in Real Domestic Saving Rates in Developing Asia
2001-07
2011-20
2021-30
Projected
PRC
31.82
Projected
30.30
Hong Kong, China
29.75
24.33
20.02
Indonesia
India
24.08
14.54
21.59
14.92
20.80
15.91
Korea, Rep. of
42.02
35.53
37.36
Malaysia
Pakistan
44.65
6.66
43.74
7.01
41.97
10.05
Philippines
Singapore
14.90
58.74
12.91
47.02
11.81
40.43
Thailand
31.31
28.59
23.53
Taipei,China
Viet Nam
25.10
16.76
20.68
19.19
15.65
15.44
Developing Asia
27.38
26.33
27.21
Economy
Fitted
31.88
Notes: Authors' calculation. Refer to the main text for explanation.
38
Future Demographic Trends
Table 3: Popu lation Ag in g in Dev elopin g Asia
The Period during which the Population
Aged 65 and Older Reaches 14 Percent
Economy
the Total Population
PRC
2020-25
Hong Kong, China
Indonesia
2010-15
2040-45
India
Korea, Rep. of
2050-55
2015-20
Malaysia
Pakistan
2040-45
After 2055
Philippines
Singapore
Thailand
2050-55
2015-20
2020-25
Taipei,China
Viet Nam
2015-20
2030-35
Japan
1990-95
Data Source: The United Nations’ (U.N.) projections available at
http://esa.un.org/unpp, and the Statistical Yearbook for Taipei, China,
available at http://www.cepd.gov.tw/encontent/m1.aspx?sNo=0000063.
39
Future Trends in Individual Countries
• Downward trend
Hong Kong, Indonesia, Korea, Malaysia,
Philippines, Singapore, Taiwan, and Thailand
• Roughly constant
China and Vietnam
• Upward trend
India and Pakistan
40
Future Trends in Developing Asia as a
Whole
• The domestic saving rate of developing Asia as
a whole is projected to remain roughly
constant for the next two decades, averaging
26.3% during the 2011-2020 period and 27.2%
during the 2021-2030 period, compared to an
actual value of 29.4% and a fitted value of
27.4% during the 2001-2007 period.
41
V. SUMMARY AND CONCLUSIONS
42
Summary (1)
• Domestic saving rates in developing Asia have
been high and rising but there has been
considerable variation from country to
country.
• Domestic saving rates in developing Asia are
determined primarily by the age structure of
the population, the level of per capita GDP,
and the degree of financial sector
development, with the latter two having a
non-linear impact thereon.
43
Summary (2)
• Domestic saving rates will decrease in the
rapidly aging countries of developing Asia over
the next two decades because the negative
impact of population aging will offset the
positive impact of rising income levels.
• Domestic saving rates will increase in the less
rapidly aging countries of developing Asia
because the negative impact of population
aging will be offset by the positive impact of
rising income levels.
44
Summary (3)
• Domestic saving rates in developing Asia as a
whole will remain roughly constant over the
next two decades.
• Shioji and Vu (2011) project that investment
rates in developing Asia will show a moderate
decline in the next two decades.
45
Summary (4)
• This implies that current account surpluses in
developing Asia may increase even further in
the next two decades, exacerbating global
imbalances even further.
• Government policies that discourage saving
(stimulate consumption) and stimulate
investment would help to reduce current
account surpluses in developing Asia and
thence global imbalances.
46
Thank you very much for your kind
attention.
47