VoxEU Column Frontiers of economic research Global economy

Why are savings so high among the young in urban China?

Household savings in China are high by international standards, and the young save as much or more than the middle-aged – a fact at odds with the standard life-cycle savings model. This column argues that neither old-age support by the middle-aged nor the one-child policy can satisfactorily explain this phenomenon. Rather, currently high housing costs and the prevalence of inter-generational shared housing are key reasons for the higher savings rates of the urban young in China.

A well-known phenomenon in contemporary China is the high personal savings rates of households compared with those in developed countries and many low-income countries. A less-studied aspect of this is the elevated savings rates of the young relative to the middle-aged, first shown by Chamon and Prasad (2010) based on urban household data covering the years 1986–2005 for ten provinces. Their finding, that savings rates for 25–40 year-olds were as high as or higher than those for the middle-aged, is at odds with the standard life-cycle savings model, which implies that relative savings rates should be low for the young, whose incomes are expected to rise over the life-cycle (Japelli and Modigliani 2005). This pattern is seen as well in the 2002 China Household Income Project data, covering 72 cities (Figure 1), and is in contrast to the life-cycle pattern of savings in the US and many European countries, for example, which exhibits the conventional inverted U-shape.

Figure 1 Household savings rate, urban male heads aged 25-65, by head’s age

Source: 2000 CHIP.

A number of studies using Chinese urban household data have addressed the issue of elevated household savings rates in China. All focus on the Chinese one-child policy as a major cause. Banerjee et al. (2012) and Choukhmane et al. (2013) suggest that the combination of smaller families induced by the policy and the tradition of family-based old-age support raise savings rates at younger ages. There are two mechanisms:

  • Young parents expect less familial support when old because they have fewer children.
  • The burden of financial assistance to the old by the younger generation when middle-aged is higher.

Wei and Zhang (2011) focus on marriage competition in China, as households with young parents seek to attract better mates for their child by providing greater financial resources at marriage. This competition is exacerbated by the well-documented effects of the one-child policy on sex ratios at birth, which creates an excess of males.

Issues with existing research on Chinese savings

There are three major problems with the existing research on household savings behaviour in China, however. The first is that the data used by all researchers do not actually represent the life-cycle pattern of savings for individuals or couples because of another important phenomenon in China – the high co-residence rates of the young with their parents. The 2002 China Household Income Project data shows the pattern of inter-generational co-residence by age for males in urban areas (Figure 2).

Figure 2 Fraction of all urban males aged 25-65 co-residing with parents or in-laws, by age

Source: 2002 CHIP.

  • Among males aged 30, half are still co-residing with at least one parent or in-law.

This is a problem for studying the life-cycle patterns of savings because the data sets provide savings at the household level, not for individuals or couples.

The reported savings rates by age are actually rates by the age of the household head, and young men co-residing with a male parent are unlikely to be assigned headship. Indeed, in the 2005 Chinese mini Census only 7% of men aged 25–45 are household heads. This age-specific censoring of savings by individuals or couples due to co-residence raises the question of whether the reported age-pattern of savings in urban China is simply a data artefact and not a true departure from the standard life-cycle model for individuals.

A second problem is that in urban areas of China today support of the old by the young is not a major phenomenon that could significantly motivate elevated savings.

  • In the 2005 Chinese mini-Census, less than 10% of men aged 65 and above report that their principal source of support is their family.

Our own data on net inter-generational financial transfers indicate that such transfers to the old are a small fraction of the incomes of the young and middle-aged. The absence of substantial old-age support is not surprising in contemporaneous urban China. Indeed, many of the current old own their homes, mostly acquired during the housing reforms at highly subsidised rates (Wang 2011). The current young, however, face unsubsidised and high housing costs, and would have to wait many years to inherit their parent’s home given the approximate 25-year age gap between parents and children. Many of the current urban old in China also have generous pensions, at replacement rates of up to 80%.

A third problem with the analyses focusing on the one-child policy is that the relatively high savings rates of the young are not unique to China.

  • The pattern of relatively high young savings rates was seen in Japan in the 1970s, as documented by Hayashi (1986).

Japan did not have a policy artificially lowering fertility. What Japan then and China today have in common are high inter-generational co-residence rates and high housing costs.

New research

In a new study (Rosenzweig and Zhang 2014), we integrate the analysis of life-cycle savings with the analysis of inter-generational co-residence. Despite the growing importance of co-residence of parents and young adult children internationally (Ruggles and Heggeness 2008), there have been few studies of the phenomenon in the economics literature, and none links co-residence to savings patterns. We also employ unique survey data we have collected that provide savings at the individual or couple level for adult twins and non-twins in five Chinese cities, as well as information on siblings and parents that is not censored by housing choices (co-residence).

We show that consideration of co-residence jointly with savings behaviour is important not only for understanding the problem of the censoring of individual savings from household survey data, but also because sharing the parental home is a potential mechanism for lowering consumption by the young, thus permitting higher savings rates. If a young adult desires to save, subsidisation of consumption via shared housing can facilitate savings. Of course, this raises the question of why the young would want to save. One reason is the high costs of housing, which would also make shared residence more desirable. Just as urban Japan in the 1980s was characterised by high housing costs and high rates of inter-generational co-residence, urban housing costs in China are also high relative to cities around the world. Rental markets in urban China are also thin, due to lack of both tenant and landlord protections, and this market failure combined with down-payment requirements for purchasing a house thus necessitates the accumulation of savings prior to leaving the parental home.

Our data enable us to identify individual (or joint married-couple) savings and not just aggregated multi-generation savings rates at the household level. These data also provide financial transfers across generations and the characteristics of all respondents’ siblings and parents that are not censored when family members choose to live apart. We show using these data that the age-selective censoring of individual savings due to co-residence accounts for a part of the life-cycle savings puzzle, but that individual savings rates are still elevated for the young relative to the middle-aged. To explain the individual patterns of life-cycle savings, we construct a multi-generation life-cycle model in which the savings and co-residence of two generations in a family are jointly determined. There are two important features of the model: the inclusion of housing services costs and a preference for privacy by the young. A key implication of the model is that many of the predictions with respect to the effects of family size and incomes on savings behaviour become ambiguous when inter-generational co-residence is an important option for families, as it is in China. We are able to generate a number of testable implications from the model for how variations in the number of siblings in the young generation affect savings rates, co-residence, and inter-generational financial transfers; how changes in housing prices affect savings and co-residence; and how changes in life-cycle incomes affect co-residence and savings.

We apply and test the model exploiting our twins survey data, using a variety of empirical methods that exploit data on twins. First, we exploit the fact that the incidence of twins in first pregnancies – a purely random event – is a good predictor of the total number of children. Our data indicate that for the cohorts of men born prior to the one-child policy those born in twins-first families have on average 0.6 more siblings. Thus we are able to estimate the effect of exogenously decreasing the number of siblings, such as induced by the one-child policy, on savings, on co-residence, and on financial transfers to parents by comparing men born into twin-first families with those men born into families without a twin-first birth. We find that an exogenous decrease in the number of siblings increases financial transfers from the young to the old, as assumed in the model and in prior studies, and that, consistent with the predictions of the model, an increase in the number of siblings increases the likelihood of co-residence. However, the number of siblings is unrelated to savings.

To investigate the mechanisms for these results, we first show that men with more siblings have significantly lower schooling attainment and incomes. The issue then is how income affects savings and the choice of co-residence. We answer these by exploiting differences between twins in a twin-pair. The advantage of looking at twin differences is that twins have the same parents, are the same age and, among identical twins, have the same genetic endowments. Across individuals, those with higher earnings also tend to have higher-earning and wealthier parents. Thus, it is difficult to isolate the effects of variations in incomes and housing choices of the young by looking across individuals, as these effects may be confounded by the effects of parental characteristics. Indeed, our model and our data indicate that wealthier parents are more likely to share their residence with their adult children. By comparing twins with different earnings we eliminate the influence of parental background. We also use the same within-twin methodology to estimate the effect of co-residence on savings, finding that, as implied by the model, savings rates are substantially higher when the young co-reside than when they do not. Thus, young men with more siblings have lower incomes, which makes them save less, but are more likely to co-reside with parents, which allows them to save more. Evidently these two effects offset each other so that the family size one is born into has no effect on savings in urban China.

Concluding remarks

All of these results, in addition to the direct evidence we adduce on the magnitudes of financial transfers across generations, suggest that in urban China neither old-age support by the middle-aged nor the one-child policy are major factors underlying the relatively high savings rates of the urban young. Rather, currently high housing costs and the prevalence of inter-generational shared housing, which itself is made more attractive when the price of housing services is high, are key reasons for the higher savings rates of the urban young in China. Indeed, our results suggest that if relaxing constraints on fertility in China increases family size, this will increase the amount of inter-generational co-residence and may further raise the relatively high savings rates of the urban young to the extent that the resulting population growth further increases housing costs.


Banerjee, Abhijit, Xin Meng, and Nancy Qian (2010), “The Life Cycle Model and Household Savings: Micro Evidence from Urban China”, Working paper, Yale University.

Chamon, Macros and Eswar Prasad (2010), “Why Are Saving Rates of Urban Households in China Rising?”, American Economic Journal: Macroeconomics, 2(1): 93–130.

Choukhmane, Taha, Nicolas Coeurdacier, and Keyu Jin (2013), “The One-Child Policy and Household Savings”, mimeo, Science Po.

Hayashi, Fumio (1986), “Why is Japan’s Saving Rate So Apparently High?”, in NBER Macroeconomics Annual 1986, Cambridge, MA: MIT Press: 147–210.

Jappelli, Tullio and Franco Modigliani (2005), “The Age-Saving Profile and the Life-Cycle Hypothesis”, in The Collected Papers of Franco Modigliani, 6, Cambridge, MA: MIT Press: 141–172.

Modigliani, Franco and Shi Larry Cao (2004), “The Chinese Saving Puzzle and the Life-Cycle Hypothesis”, Journal of Economic Literature, 42(1): 145–170.

Rosenzweig, Mark and Junsen Zhang (2014), “Co-residence, Life-Cycle Savings and Inter-generational Support in Urban China”, NBER Working Paper 20057.

Ruggles, Steven and Misty Heggeness (2008), “Inter-generational Co-residence in Developing Countries”, Population and Development Review, 34(2): 253–281.

Wang, Shing-Yi (2011), “State Misallocation and Housing Prices: Theory and Evidence from China”, American Economic Review, 101(5): 2081–2107.

Wei, Shang-Jin and Xiaobo Zhang (2011), “The Competitive Saving Motive: Evidence from Rising Sex Ratios and Savings Rates in China”, Journal of Political Economy, 119(3): 511–564.

4,619 Reads