Modigliani - The Chinese Saving Puzzle(2004) - edited

From this paper:

"In the Chinese cultural tradition, the younger generation is supposed to take care of the elder members of the family, while the elders will bequeath the house and other assets to their children. In other words, an economic unit is the extended family rather than the nuclear
family. Under such a system, a child is an effective substitute for life-cycle saving.
Consequently, when strict birth-control measures came into effect in the 1970s, the accumulation of life-cycle (tangible) assets gained in importance as a substitute for
children. One may also argue that even if the Chinese birth-control policy did not
occur, the secular trends of 1) more nuclear families; 2) migration of families
from their ancestral homes; and 3) less loyalty to elders would have had the effect of
reducing the role of children as a substitute for saving, anyway.
(...)
By the early 90s, the Chinese personal saving rate had reached a remarkable level of nearly 30
percent with a peak of over 33 percent. This occurred despite the fact that, even with the high growth rate, the per-capita income remained one of the lowest in the world. These saving rates are stunningly high in comparison with those of the United States, one of the world’s richest nations. During those same years, the personal saving rate in the United States was 7.6 percent;
and even the “private” saving rate, which is the sum of personal saving and corporate saving (profit retention), rises to only 10 percent. Since then the saving rate has
slipped further with the personal down to 3 percent and the private rate down to 5 percent—
though these low figures may reflect partly a transitory response to the boom in asset values.

When lay people are confronted with these figures, their usual reaction, after rubbing
their eyes, is to attribute the huge gap to obvious differences in upbringing and education. This thinking reflects cultural-ethical values attributed to personal thriftiness and risk-taking in different cultures. But, if the reader has found the analysis of this paper persuasive, he should understand that that type of explanation is fundamentally baseless. And the simplest proof is found in this very paper that for a long stretch of very recent time, 1958–1975, the Chinese
saving rate was quite low, around 5.3 percent or lower than in the United States.
What then is the explanation?
The key to the puzzle, we suggest, is provided by the LCH(*) and its implication that the major systematic determinant of the rate of private saving is to be found in the rate of growth of income and the demographic structure of the economy, while per-capita income, the traditional and commonsensical explanation counts little, if any. According to this model the extraordinary behavior of the Chinese saving ratio is the result of two nearly coincidental sharp turns in two key policies. The first is the movement initiated in the late 1970s toward a market-oriented economy, which, along with a number of very special characteristics of China’s society and labor force made possible an explosive growth pattern
such as was never seen before. With this development the growth rate jumps from a
more or less stable rate of 4 percent to a gradually rising (accelerating) pattern
reaching some 12 percent but 25 years later. The second turn regards demographic
policies. Until the 1970s the Chinese government was not seriously concerned with
population growth, and in fact for a while under Mao there was an endeavor to
encourage births. But eventually the view prevailed that to improve the economic wellbeing
of the Chinese population, it was essential to control population growth, and
the new policy was announced and strictly enforced to limit the number of children per
family (just one in the cities). As noted, this had a double profound effect on the saving
ratio. The first was a drastic decline in the ratio of people under fifteen years to working
population from 0.96 in the mid-70s to 0.41 at the turn of the century. The second
was to undermine the traditional role of the family in providing old-age support to the
parents by the children, thus encouraging provisions through individual accumulation.

According to our estimates each of these developments contributed equally more than ten basis points to the rise in the saving rate of some thirty basis points (from 3 percent to 33 percent) with the remainder largely accounted for by the spurious effect of inflation (five basis points).
As a further demonstration that the prodigious saving rate reached by the mid 90s does
not reflect “ethnic” or cultural characteristics of China, we report illustrative episodes for other countries and periods.
Japan had an exceptionally high growth rate—though not quite matching that of China—and a quite favorable population structure that had a saving rate that nearly matched China’s. One might be tempted to say that, after all, China and Japan share the heritage of the East. Unfortunately, such a simple explanation is refuted. In the 1960s, Italy saved even a larger fraction of its income than Japan. Yet, one cannot even use the excuse of the “Protestant ethic” since Italy is a solidly Catholic country and many think of it as the “paese della dolce vita”!
The explanation can again be found in a growth rate not as high as the other two countries but still well above average (row 7) and a very favorable M/P that reflected the sharp decline in population growth, to which one can probably add the substantial loss of its tangible and intangible wealth during World War II. And a similar story can be repeated for France with a fairly high growth rate of 5 percent, saving 19 percent of its income, and of Portugal with a somewhat higher growth rate and a saving ratio of nearly 20 percent.
In OECD countries during the 60s, the growth rate peaked at 4.9 percent and the saving rate also peaked for most countries with an average of 14.8 percent. Both are impressive by today’s standards. By the 1980s, the average growth rate had declined to below 3 percent and the saving rate to 14.4 percent. The inflation-adjusted saving rate had declined even more, from 14.8 percent to 12.3 percent.

(*)Life-Cycle Hypothesis: lifetime consumption profile expected to be essentially flat, with people borrowing against future earnings during their early study and working life when income is low, saving greatly during their most productive working years and consuming saved assets during retirement.