international comparison of saving ratios by occupational...
TRANSCRIPT
INTERNATIONAL COMPARISON OF SAVlNG RATIOS BY OCCUPATIONAL GROUPS (ID*
By TOSHIYUKI MIZOGUCH1**
I11. Saving Ratios of Self-Employed Families
It hus been an imporLant topic in the analysis of the consumption function to examine
the saving ratios in self-employed families. Some papers proved that the average propensity
to consume was lower in these families than in workers' families.13 Interesting assertions
have also been made by Japanese economists, depending on the experience of Western de-
veloped countries. When we investigate the national accounts statistics published by the
United Nations, the percentage of wage-type income in total personal income (W-Pratio) is
lower in Japan than in most Western developed countries, as is shown in Table 12. If the saving
TABLE 12. COMPARISON ol* RA1'ros or* WAGE-11'YI'E INCOlvlE IN PERSO*N!~L INCOlvlE
Source: U.N.. Yearbook of National Accounts Statistics, 1965.
* The first part of this paper was published in this Journal, February, 1967.
** Assistant Professor (Jokyo~ju), Institute of Economic Research.
18 For example see Klein. L. R., and Margolis. J., "Statistical Studies of Un mcorporated Buslnesses "
Reviezv of Econo'nics and Statistics, February, 1954.
28 HITOTSUBASHI JOURNAL OF EcoNoMlcs [June ratios are similar among the countries by occupational groups, the lower percentage of wage-
type income induces the higher personal saving ratio, and this can explain the high personal
saving ratio in Japan.14 In order to support this assumption, we must compare the saving ratios of self-employed families among the countries. In this examination, it is desirable to
divide the self-employed families into at least two sub-groups: the farm families and the non-
farm self-employed families, because the saving behavior is considered to be different between
these groups and the number of households in the latter group is too large to be neglected,
even in Japan in recent years.
First let us compare the saving ratios of farm families. The international comparison is
more difficult in this group than in the case of workers' families because of the scarcity of
statistical materials. There are some family budget surveys on farm families, but most of
them give no information on their incomes. We can find only three family budget surveys
on farm families for our purpose: in Finland, the Ryukyu and the United States.15 Two
family saving surveys in Sweden and India can be used as additional information. In addi-
tion, special surveys on farm families, called the Fartn Household Economic Survey (NOKA
KEIZAI CHOSA in Japanese), can be found in Japan, Taiwan and South Korea. These surveys attempt to compile comprehensive figures on activities of farm families; farm produc-
tions, investments, incomes, consumptions, assets and liabilities, etc. Therefore, we can get
statistics on the saving ratios of farm families in eigllt countries, as is shown in Table 13.
Another difficulty arises from the different definitions of farm households. As the scales of
farms or the production techniques are very different among the countries, it is difficult to
TABLE 13. STATISTICAL MATERIALS FOR THE ANALYSIS OF SAVlNG RATIOS OF FARM FAMILIES
Note: 1) The sample sizes shown in this table are those in recent publications and they have been fluctuating year by year, though the ranges are not too large.
2) The letters in the column "type of survey" indicate as follows; FH: the farm house-hold economic survey. FB: family budget survey, S: family saving survey, BS: the surveys with balance sheets shown in Table 1, WBS: the surveys without the balance sheets.
3) All surveys, except the earlier surveys in China (Taiwan) and Korea (South), adopt the random sampling technique.
4) As for the data sources, see the reference lists shown in the final part of this paper.
14 See, Miyohei Shinohara, Growth and Cycles in the Japanese Economy, Kinokuniya, 1962. 15 There are some family budget surveys with income information including farm families, but without
occupationa] classifications. These surveys are found in the Phillippines, Thailand and in some African
countries.
1967] INTERNATIONAL COMPARISON OF SAVING RATIOS BY OCCUPATIONAL GROUPS (n) 29
define the self-employed farm families. In this paper, although employed farm labourers are
excluded, the other definitions in each country are used without any adjustments. Some reserva-
tions are necessary, therefore, in the following conclusions. We must also consider the definition
of income and saving in each survey. As the amount of depreciation of business equipment
is not negligible among farm families, the saving ratios would be different in the gross concept
or in the net concept defined in Table 3. All surveys, except the Farm Economic Household
Survey in Japan, adopt one of the two definitions, and it is very difiicult to adjust them to
the same concept. In the following analysis, some tentative adjustments will be indicated,
but some errors are inevitable.
Let us now proceed to the international comparison of saving ratios of farm families.
The saving ratios with brackets are calculated by supposing a 5% depreciation rate for farm
equipment,16 and the others are referred from the orig'inal data. Investigating this table, the
saving ratio of Japanese farm families is not necessarily higher than in the other countries.
TABLE 14. SAVlNG RATIOS OF FARM FAMILIES
Note: 1. a) adjusted for depreciation on residential houses and consumer durables, b)
excludlng the increase of cash.
Further, a comparison of Table 14 with Table 5 shows an interesting result. Generally speak-
ing, the saving ratios are higher in the farm families than in workers' families, though the
incomes of farm families are lower than those of workers' families. A noticeable exception
can be found in the Japanese case. The saving ratios of farm families are significantly lower
than in workers' families in this country. In view of this, it may prove interesting to de-
cornpose the saving ratios. As the data for such a purpose are not too abundant, it is
dangerous to assert decisive conclusions, but a comparison of figures in Table 15 with those
in Table 9 does give some suggestions. The higher saving ratios in farm families originate
mainly from the investment ratios in farm families. In contrast with farm families in the
other countries, the net business investment ratio is negative in Japan, and this low ratio
explains why the saving ratios are lower in farm families than in workers' families in Japan.
This fact may be partially explained by the higher depreciation ratios adopted in the Farrn
Household Economic Survry in Japan. That is, the average depreciation ratiolT in 1961 was
16 A 5~/o depreciation rate is adopted in the family budget surveys in the United States in 1961.
17 This ratio is calculated by dividing the total depreciation values by the book values of the farm
equipment.
30
TABLE
IlITOTSUBASnl JOURNAL OF EcoNOMICS
15. SAVING RATIOS BY THEIR CoMPONEN1'S OF FARM FAMILIES
[ June
Per Capita Personal Income ('60 U.S. $)
L{quid Type Saving Ratio (~)
Life Insurance Premium Ratio (~6)
Investment Ratio (Personal Use) (~)
Business Investment Ratio (~;~)
Decrease of Liabilities' Ratio (%)
} }
Gross
Japan 1961
315
16. 2
6.
-2.
7
4
Concept
Sweden 1 958
632 1. 3 O. 7 1. 7
28. 3
-12. 7
about 10% and this is rather higher than the ratio in the United States. But a more funda-
mental reason should be found in the recent tendencies in the economic behaviour of Japanese
farm families. The rapid growth of the economy in post-war Japan has increased the income
differentials between the urban sector and the rural sector. Therefore. Japanese farmers tend
to increase their incomes by working for enterprises in the urban sector, and the farm labour
has been supplied by their wives or aged labourers, whose wdges in the urban sector are not
too high yet. This fact can be investigated numerically in the time-series analysis of the
Fal-m Housel,old Economic Survey in Japan. For example, the percentage of farm income in total income in farm households has been rapidly decreasing in recent years in Japan, as
is shown in Figure 3. Since the scale of farms in Japan is generally small, it is difficult to
FIG. 3. TIME-SERIES CHANGES OF FARM INCOME IN TOTAL INCOME OF JAPANESE FARM FAMILIES
(%) 100
90
80
70
60
50
40
30
o
Non-Farm Income
Farm Income
'52 '5 5
'6 o ( Year )
Note: As the survey method of Far,,1 Household Economic Survcy in Japan was revised twice, we cannot use the original data for a time-series.
This figure is written by using a special link technique calculated by the
Statistical Section. Institute of Economic Research. Hitotsubashi University.
1967] lNTERNATIONAL COMPARISON OF SAVING RATIOS BY OCCUPAT10NAL GROUPS(II) 31
increase their productivity by business investment as rapidly as the growth of real income in
the urban sector.The incentive to invest ln farming,therefore,is not very great.In addi・
tion,some farmers decreased their land by selling to the urban sector because of the scarcity
of farm labour.The rise of land prices,which was mentioned in Section II,accelerates this
tendency.Amorecomprehensivemodelwouldbenecessarytoexplainthisphenomenon,because the economic behaviour of farm families in recent years in Japan is very complicated.
But it may be safely said that the saving ratios in Japanese farm families are not higher than
those in the other countries and that the business investment ratios in Japan are significantly
lower than in the Westem developed countries.
Further,some investigations of the factors explaining the differences of saving ratios in
farm families among the countries should be made.One approach may be to make the ngure
of time・series changes of saving ratios of farm families corresponding to Figure1,The result
shown in Figure4seems to be similar to that shown in Figure1,though the number of
FIG.4,TIME-sERIEs CHANGEs oF SAvING RATlos oF FARM FAMILIEs (A) 51/ツ (B) 52/ツ
)【」
%2
20
15
10
5
0
(%) 穏糊61
,63 25
窟 噂1
摩 2。 ’62 ,62 Sweden ,61 ’41 互呂窪魯嘉9 ’58煽161 15
曙 ロ き ’ξ
← ・61 少52,57 10
郵9 5 躍 娼
●召 P61
←β .
In-ia.
’638 Fin1an己,59 &’61 ,63
著, 0
50 100 .5001,0002,000 50 100 500 1・0002・000 ,63 Finland,58 ’54 (PerCapitalnc。me,’60U.S.$) ’54(PerCapitalnc。me,,60U,S.$)
Note; L ×一×:in gross conceptン ・一・;in net concept
2.As for the Japanese case,see the note in Figure3。
samples in this complete Hgure ls not enough. The next examlnation is to correlate the
saving ratios with factors adopted in the analysis in Table6,but the result is not too suc・
cessfuL Apartofthismaybecausedbythedecisionprocessofbusinessinvestment.Asis shown in Table15,the share of business investment is rdatively large in the total saving
of farm families.This mtio is considered to depend on the production techniques and on the
scales of farms,so the total saving ratios would not necessarily be explained by the income
conditions of farm famihes. In this respect,it is important to point out that the liquid・type
saving ratios are correlated,to some extent,with the growth rates of real income of agricul-
ture and forestry,
Now Iet us proceed to the intemational comparison of the saving ratios of the non・farm
32 HITOTSUBASHI JOURNAL OF ECONOMICS [June self-employed families. This family group is sub-divided again into two sub-groups; i.e., the
non-farm entrepreneur families and professional families. However, there is not very much
statistical materials on these groups. We can find three family budget surveys on these groups
in the Ryukyu, the United Kingdom, and the United States, as well as the four family saving
surveys in India. Japan, the United Kingdom and the United States. These surveys are listed
in Table 16. In examining the saving ratios of these groups, there are difficulties similar to
TABLE 16. DATA FOR THE ANALYSIS OF SAVING RATIOS OF
NoN-FARM SELF・EMPLOYED FAMILIES
Note: 1. As for the letters in the column "Type of Survey" see note m Table 13 2. The letters in the column "Scope of Survey" are as follows: E; Non-farm self-employed
entrepreneurs'. P; professionals', SE; Non-farm self-employed families.
those encountered in the case of farm families; i.e, the differences between the definitions of
occupations and between the net concepts and the gross concepts.
First, Iet us compare the saving ratios. Though the number of countries in Table 17 is
not enough to get any conclusive results, some suggestions can be found. It should be noted
that the saving ratios in Japan and the Ryukyu are much higher than in the other countries.
TABLE 17. SAVlNG RATIOS 017 NoN-FARM SELF-EMPLOYED FAMILIES
Non-f arm Self・employed Entrepreneurs
India
Japan
Ryukyu Sweden U. K.
U. S. A.
1960
1963
1964
1958b)
1962")
19eo/,61
93
(277)
(1, 235)
l, 140
2, 340
5. 4
(24. 8)
(11. 8)
10. O
11. l
5. 5
(31 . 4)
(22. 3)
17. O
22. 3
326
249
1, 315
29.
32.
19.
9c) d]
2 3c]
36 .
34.
30.
5c) d]
3 3c]
Prof essionals
Japan
Ryukyu U. S. A.
1963
1964
1960/'61 2, 540 8. 8
20. 5
522
737
31.
59 .
Oc) d)
7c)
35.
63.
lc] d)
2c)
Note: 1, a) Including the Excluding the increase of cash,
2. The figures in ( ) are
professionals, b) Including the families without
d) Excluding the sales of business buildings .
estimated by using information other than that
occupations, c)
in the surveys.
1967] INTERNATIONAL colt,IPltRrsoN OF SAVlNG RATIOS BY OCCUPATIONAL GROUPS (n) 33
Some papers have suggested that the saving ratios in the non-farm self-employed families may
be very high in Japan, though these papers do not support their suggestions with rigid inter-
national comparisons.18 The result in Table 17 supports these previous suggestions; that is,
the saving ratios of self-employed entrepreneurs are much higher in Japan and the Ryukyu
than in the other countries. This fact may be related to the high speed growth of their
income in Japan and the Ryukyu. When we calculate the average annual growth rates of
real income in commerce in the four years before the surveys, the values in the two countries
are above 10%, but they are under 5% in the other countries. The higher levels of the
saving ratios of the professional families may be explained by same reason, but an additional
factor should be mentioned. A distinguishing characteristic of Table 18 is the surprisingly
TABLE 18. lr~AC'rORs DETERMINING Tl~E SAVING RATIOS Or
NoN-FARM SELF-EMPLOYED ENTREPRENEURS (Unit: ~~)
Note: 1. As for the definitions of variables, refer to Table 6.
high saving_ ratios of professional families in the Ryukyu, and these figures may be explained f i we compare their incomes with those of workers famllres The number of persons belong-
ing to the porfessional category is re]atively large in the Western developed countries, especi-
ally in the United States. But in the less developed countries, such as the Ryukyu, only a
limited number of people engage in the professions and the relative difference between their
income levels and those of other occupational groups tends to be greater than in developed
countries. The same tendency can be also found in Japanese professional families, though
the income differences are not so significant as in the Ryukyu.
It is desirable to attempt again to decompose the saving ratios of these groups, The
figures in Table 19 suggest that the liquid-type saving ratios in Japan are very high, but the bisiness investment ratios as well as the investment ratios for personal uses are not too large.
This result is very similar to the special characteristics of Japanese workers' families, and
may be related to the high speed growth of the Japanese economy. However, in the case of
self-employed families, the financial conditions should also be considered as a supplemental
factor, It is commonly said that the high speed growth in post-war Japan was supported by
the increase of bank loans to large enterprises. In fact, the demand for liquid-type capital
is quite considerable large in Japanese enterprises in order to supply the funds for large scale
investments in equipments. As a result, the loan funds for small merchants or artisans are
rs See, for example, M. Shinohara, Growth and Cycles in the Japanese Economy, op., cit., Ryutaro Komiya, Post-War Japanese Economic Growth, University of California Press, 1966, Ch. 8.
34 HITOTSUBASHI JOURNAL OF EcoNoMlcs
TABLE 19. SAVING RATIOS BY THEIR COMPONENTS OF NoN-FARM SELF-EMPLOYED FAMILIES
[ J u ne
(Unit: ~o)
Note: 1. a) including the increase of cash, b) excluding the increase of bonds, c) including
the increase of bonds, d) the depreciation of business buildings is not deducted.
very restricted, and it is difficult for them to get loans or credits to their desired amounts.
Further, the interest rates for them are said to be very high compared with the other countries.
In some cases, they are even asked to save some amount of tirne-deposits as a guarantee for
loans from banks, (this strange financial system is called BUZUMI and RYODATE in Japanese).
These financial restrictions tend to induce an increase in the level of the "desired stocks" of
saving among Japanese merchants and artisans. As is shown in Table 20, the saving-stock-
TABLE 20. SAvlNG STOCKS INCOME RATIOS OF SELF EMPLOYED FAMILIES (Unit: ole)
Note: 1. F: farm famllies, NF, non-farm self-employed families. A: all families.
income ratios in liquid-type saving are not lower in Japan than in the other countries. But,
remembering the financial restrictions mentioned above, the stocks of liquid-type savings are
rather lower than the desired ratios in Japan. Further, the higher rate of income growth
requires higher flow saving ratios in order to keep the saving-stock-income ratios constant.
These assertions would not be enough to explain the significantly high levels of saving ratios
of Japanese non-farm self-employed families, but the writer thinks that the problem will be
solved if we examine the reason why the liquid type saving ratios are very high in the non-
farm self-employed families as well as in workers' families in Japan.
Returning now to the initial topic, Iet us examine whether higher personal saving ratios
in Japan can be explained by the large share of non-wage-type income in total personal in-
come. As the saving ratios are higher in non-farm self-employed families than in workers'
1967] INTERNATIONAL COMPARISON OF SAvlNG RATIOS BY OCCUPATIONAL GROUPS (n) 35
families in Japan, the initial assumption is supported by the data for these groups. But the
results for farm families have the opposite implication; that is, the saving ratios of the farm
families are significantly lower than in the workers' families in this country. As the number
of farm families is much larger than that of the non-farm self-employed families, the average
result gives a pessimistic answer for the assertion. When we calculate the saving ratios of
self-employed families, the Japanese values are not too much higher than those in Sweden and
the United States, as was indicated in Table 17. Therefore, it is dangerous to assert some
hypotheses by using the national accounts statistics without any detailed examinations.
TABLE 21. COMPARISON OF SELF-EMPLOYED FAMILIES IN JAPAN, SWEDEN AND THE UNITED STATES
(Unit: ~)
Note: l. Average saving ratios of all self-employed families are calculat8d by the ratios of
weighted average of savings to weighted averages of income with distribution of the number of households.
IV. Saving Ratios of Families without Occupations ard
the Life-Cycle Approaches
Finally, Iet us examine the saving ratio of families without occupations. These family
groups consist of two different sub-groups; i,e., retired and unemployed families. The latter
sub-group is important in the analysis of saving ratios in underdeveloped countries. Their
incomes are generally low and saving ratios are negative. As unemployment increases in
economic recessions, the personal saving ratios may fluctuate violently in these countries.
Discussions of the saving ratios of retired families have been developed in the Western developed
countries, as a part of life-cycle theories. The weights of retired people in the developed coun-
tries are not negligible, and the large share of the group is considered a cause of decreasing
the average personal saving ratios in these countries. From this point of view, the recent
situation in Japan is interesting; i.e., the number of unemployed families is negligible but the
weight of the retired families is also small, though not negligible. Table 22 shows that the
weight of families without occupations is much lower than in the Western developed countries.
This fact can be explained as follows: (1) the rate of unemployment of young families is
10w in recent years in Japan, (9-) the percentage of aged families is lower than in the Western
developed countries, (3) aged persons find new employment with a lower wage rate or salary
after retiring from their main jobs, (4) some aged persons live together with their sons and
such families are classified by the occupations of their sons. As for (1), there are few objec-
tions among Japanese economists. Therefore, the life cycle analysis for the factors from (2)
36
TABLE 22. WEIGHT
HITOTSUBASHI JOURNAL OF EcoNoMlcs
OF FAMILIES WITHOUT OCCUPATIONS IN TOTAL FAMILIES
[ June
Note: l. This tab]e is made by using the sample distribution in the surveys in the whole country.
to (4) will be the main topics in this sectlon. The analysis of factor (2) in particular has ,a
special meaning in Japan, because some papers have asserted that this factor is a cause of
the higher saving ratios in Japan. There are two sorts of statistical materials for our purpose; i.e., data for families without
occupations and head-of-household age-group figures, whose lists are shown in Table 23. A
TABI_E 23. LISTS OF SURVEYS OF SAVING RATIOS OF FAMILIES ¥vITI{OUT OCCUPATIONS AND OF AGE GROUPS OF FAMILY HEADS
Country Sample Size Type of Survey Scope of Survey
Families witbout Occupations
Belgium 1961
India 19(X)
Japan 1959-
Netherlands 1950
Ryukyu 1961-
United Kingdom 1950-'56
United States 1950, 1960
FB (BS)
S (WBS) S (WBS) FB (WBS) FB (BS)
S (WBS) FB (BS)
R. in the whole country
WF and R. in the urban area
WF in the whole country
R. in the whole country
WF in the whole country
R. in the urban area
WF and R in the whole country
Age Groups of Family Heads
Japan 1959-
Sweden 1957, 1958
United Kingdom 1950-'or6
United S*tates, 1950, 1960
S (WBS) S (WBS) S (WBS) FB (BS)
NF in the whole country
A in the whole country
NF in the whole country
NF in the whole country
Note I the symbols m the "type of survey" column are defined in Table 13. 2. the letters in the "scope of survey" column are as follows: R.; retired fami]ies, WF; families without oc-
cupatlons, A; all families, NF; non-farm families.
comparison of saving ratios in the families without occupations indicates that the ratios is
much higher in Japan than in the other countries. Consistent results can be found in the
life cycle changes of saving ratios in Japan. As is shown in Figure 5, (A), the saving ratios
show an upward trend in the young and middle ages as the age of the household heads in-
creases, but the ratios begin to decrease after 50 years old in the Western developed countries.
INTERNATIONAL COMPARISON OF SAVING RATIOS BY OCCUPATIONAL GROUPS (ID 37 1967]
TABLE 24. COMPARISON OF SAVING RATIOS OF FAMILIES WITHOUT OCCUPATIONS
Note: 1, a) excluding the increase of cash, b) including depreciation of private house and
of consumer durables.
But the life-cycle pattern of saving ratios shows a rather peculiar form in Japan. That is,
the saving ratios are very high even in aged families. This fact was first pointed out by Mr.
Kanamori.19 He explained this as a result of the special system of wages and salaries in
FIG. 5. LIFE-CYCLE CFIANGES OF SAVING RATIOS AND RELATIVE INCOME
(A) Saving Ratios (B) Relative Income (%) 34
32
30
28
26
24
22
20
18
16
14
12
10
8 6
4
2 O
.~ -2 ~; eE: -4
bO~ .F -~ ~ ~ -8 uQ
- O
s ~ ~,,,
/'¥
/~ '¥
/~v ~ '//~-e /~>/S"/ ~1 ~ ~~:-
v _ ¥,./ ¥ /¥ _- -'
p -- V r/ 1! '?rl~/~ +'+ / ~l fr
~ .-// l'¥' (f// /~i ' '¥" '-)~ '~¥
~l fl !l
'I /
**:¥. ¥,'¥
¥¥
140 (%
130
120
l 10
g, ~ o* 100 ,~ 4) uD ~ 90 ~ ? < ~ 80 " ~ o ~ E: - 70 f~ F,
e * (:) 60
~ o ~ 50 0~ L
~
//'i ,, y~./. '
-7rl //' 'l
/<~)~j/ / / /
jit
ji Ahr::~t,
l '/j ¥¥¥<¥( ! . ~ '~ ~
i'¥ ! //1 ¥t ' j~'¥il)If-'~ j '11
.. ~ *- lc"">" y.'/ 'ilndra ¥.¥t
' .(It!~c2LJ ¥~
;'/' f ~~ /~~~
/ /
20 25 30 35 40 45 50 55 60 65 70 20 25 30 35 40 45 50 55 60 65 70 Age of Family Head Age Qf Family Head
Note: l. The relative incomes are defined as the ratios of the per
capita incomes of each head・of-household age group to total per capita income.
19 Kanamori, Hisao, "Why the saving ratios in Japan are too high ?" Monthly Review of the Planning Agency (in Japanese), 1961.
Economic
38 HITOTSUBASHI JOURNAL OF ECONOMICS [June Japan;20 i.e., Japanese workers are guaranteed by this system to increase their income as they
become older. The latter assertions should, however, be carefully re-examined because the
relative income changes among head-of-household age groups, are similar among the countries
as is shown in Figure 5, (B). In order to propose another assumption, it is useful to decompose the saving ratios of
aged families in Japan. The figures in Table 25 tell us that investments in their own houses
TABLE 25. SAVlNG RATIOS BY THElR COMPONENTS OF AGED FAMILIES IN JAPAN
and in stocks or bonds are relatively large in aged families in Japan. The same tendency
can also be found m the families without occupations in Japan, and is considered a distinguish-
ing characteristic, compared with the saving behavior in the 'other countries. The writer
thinks that the retirement fund system in Japan is the main cause of this tendency. Though
the retirement pension fund system has been developing in recent years in Japan, most retire-
ment money is usually paid at the time of retirement. As this amount of money is relatively
large in comparison with their usual money income, most retired people tend to save it for
future living expenses. As residence conditions are especialy inadequate, as was pointed out
in Section II, many retired people tend to invest it in their own houses.
The above finding could support the view that the higher saving ratios of families without
occupations is also a supplemental factor in increasing the total personal saving ratio in Japan.
In addition, the lower percentage of aged families should be examined. As was discussed in
the begining of this section, the aged groups tend to work in their second jobs and this
phenomenon would be related to the incompleteness of the social security system in Japan.
Most persons of these groups lost their assets in the Second World War and the violent in-
flation after the war, and the pension fund system does not provide enough to maintain their
living expenses, even if they received some amounts of retirement money.21 In fact, the lower
income classes in Japanese family budget surveys consist mainly of young workers and old
workers. Further, we should examine factor (4), mentioned in the begining of this section.
l'es but this In pre-war Japan, the aged people used to be included in the eldest sons' fami I ,
custom has been rapidly vanishing since the end of the war. However, this custom cannot be said to be negligible even in the 1900's. In fact, when we compare the men's distribution
by ages in the population statistics on individual bases with the numbers of family budget
statistics on family bases, the shares of families with aged family heads are rather lower in
20 wages in Japan generally increase as the years of work are longer, even if the sort of work does
not change. This system is called the wage system by order of age in Japan. 21 The description in the main sentence does not necessarily mean that old people cannot maintain their
minimum standard of living. As the growth rate of personal income in post-war Japan has been very hrgh the consumptlon patterns have changed remarkably. In order to follow these changes, the amounts
of stocks of savings or pension funds are not sufficient for old people.
1967] INTERNATIONAL COMPARISON OF SAVlNG RATIOS BY OCCUPATIONAL GROUPS (n) 39
Japan in comparison with the corresponding results in the United States. Since the incomes
of aged persons who live together with their sons are considered to be low, this remaining
custom may have a , positive effect in increasing the saving ratios. But this effect may be
small in recent years in Japan.
V. Final Remarks
The international comparisons shown in the first part of this paper as well as in the
second part, made clear some distinguishing characteristics in Japanese saving behaviour.
First, the saving ratios of the workers' families are higher in Japan than in the Western de-
veloped countries. This higher saving ratio may be explained mainly by the higher growth
rate of real wages in Japan. In addition to the increase of real income of workers' families,
the increase of bonus vers, to total-income-ratios should also be mentioned. That is, Japanese
workers and salaried employees receive their bonus payments, at least twice a year and this
share in total income has been increasing since 1951. The relationship between the increase
of bonus payments and the rise of saving ratios should be examined in detail in another paper
in the future. But it appears certain to the writer that the relatively higher rates of growth
of bonus payrnents accelerated the rise of saving ratios in Japanese workers' families. The
higher saving ratios of non-farm self-employed families should be mentioned, but the saving
ratios of farm families are not too high. The latter fact negates the assertion that the higher
personal saving ratios in Japan can be explained by the lower percentage of wage-type in-
come in total personal income.
Another distinguishing characteristic, which has not been pointed out by Japanese econo-
mists, is that the liquid-type saving ratios of workers' families as well as of non-farm self-
employed families are higher than in the other countries and these differences can explain
those in the total personal savlng ratios. This characteristic, the writer believes, may be the
key to explain the higher personal saving ratios in Japan. Some tentative assertions have
been proposed in this paper; that is, the scarcity of liquid-type capital, compared with the
high growth rate of the Japanese economy, have induced the higher desired levels of saving-
stocks-income ratios in these two groups. But more detailed analyses would be required in
order to open the door for the secret room of the surprisingly high saving ratios in Japan.
The effects of special life cycle patterns on the saving ratios have also been examined.
These results are interesting because of the peculiarities of the life cycle pattern. But this
effect would remain only as a supplemental factor in explaining the highar saving ratio. There
are other assumptions to explain the fact. According to Komiya's review,22 these assumptions
are classified into 8 groups. The main part of these groups has been partially checked in
this paper, by making international comparisons. But there remain some assumptions. The main assumptions which have not been examined in this paper are as follows. (1) The
income distribution in Japan is more unequal than in the Western developed countries, and
this inequality induces the higher personal saving ratios in Japan, (2) the weight of property
incomes in total income is large in Japan, (3) Japanese people have a custom to save more,
Among these assumptions, we cannot check (3) numerically. As for (1) and (2), Komiya
22 Komiya. R.. Post-War Japa'lese Economic Growth, op. c~t
40 HITOTSUBASHI JOURNAL OF ECONOMICS
gives negative
results, though
suggestions.
a more rigid
The writer's preliminary examination is consistent with
assertion must be postponed until future examinations.
Komiya's
Reference Data (for (II) part only)
China (Taiwan): Department of Agriculture and Forestry, Provincial Government of Taiwan , Report
on Investigotion of Far'n F"conomy of Main Croppillg Far'n Fa'nily, 1954, in 'raiwan, 1956.
Department of Agriculture and Forestry, Provincial Government of Taiwan, Repo"I of Far'n Record-Keeping Families in Taiwan, 1963, 1964-.
Finland: Socialisa Erikostkimuksia. Rural Consutnption Investl~ation, 1959-1960, 1962.
India: National Council of Applied Economic Research. Urban Income and Saving, 1962. National Council of Applied Economic Research, All India Rural Household Econo'nic Surve:y,
1965.
Japan: Bureau of Statistics. Prime Minister's Office, Family Saving Survey, 1959-, 1961-. Department of Agriculture and Forestry, No~ka Keizai Chosa Hokoku, 1950-, 1952-.
Korea (South): Economic Planning Board. Korea Statistical Yearbook, 1960-, 1960-. Ryukyu: Statistical Section, Planning Board. Ryukyu 'ro~hei Gepp~, 1958-, 1958-.
Sweden: Meddelanden fran Konjunkturinstitutet, Hushdllens Sprande dr 1957, 1958, 1964-1965. United Kingdom: Oxford University, Bulletin r]f OxJb,'d University, Institute of Statistics, 1951-
1956, 1951-1956. United Kingdom: Ministry of Labour, Fa'nily Expenditure Survey, 1961-, 1963-. United States of America: Bureau of Labour Statistics, "Consumer Expenditure and Income, Urban
United States, 1960-1961", B. L. S. Report, No. 237-238, 1964.
Department of Agriculture, "Consumer Expenditure and Income Rural Farm Populatlon United States", 1964.
B. L. S., Fa'nily Spending and Saving in Warti,ne, 1954.
Federal Reserve System, Federal Reserve Bulletin, Vol. 36-42, 1950-1956.
Wharton School of Finance and Commerce, Study of Consu'ne" Expenditure and Incon'e Survey, University of Pennsylvania, 1958.
General: International Labour Office, Year Book of Labour Statistics, 1955-, 1955-. Genera]: International Labour Office, Bulletin on Fa'nily Budget Survey, 1950-1960, 1961.
United Nations: De'nographic Yearbook, 1955-, 1955-. United Nations: Yearbook of National Accounts Statistics, 1955-, 1955-.